Design and Synthesis of Inhibitors of dTDP-D-glucose 4,6~
dehydratase (RmlB), an Enzyme Required for dTDP-L-Rhamnose
Production in M.Tuberculosis
by
Neena Sujata Kadaba
S.B. Chemistry
Massachusetts Institute of Technology, 2002
SUBMITFED TO THE BIOLOGICAL ENGINEERING DIVISION
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE IN MOLECULAR AND SYSTEMS TOXICOLOGY
AT THE
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
JUNE 2003
© 2003 Neena S. Kadaba. All rights reserved.
The author hereby grants to MIT permission to reproduce and to distribute publicly
paper and electronic copies of this thesis document in whole or in part.
Signature of A uthor.............................................................................
V ....................................
Neena S. Kadaba
Biological Engineering Division
May 9, 2003
Certified By...................................................
**
..............
. ........................................
\ jjJohn M. Essigmann
Professorof Chemistry and Toxicology
Thesis Advisor
A ccepted By ......................................................................
2 ...................................................
Ram Sasisekharan
Chair, Biological Engineering Graduate Committee
Design and Synthesis of Inhibitors of dTDP-D-glucose 4,6dehydratase (RmlB), an Enzyme Required for dTDP-L-Rhamnose
Production in M.Tuberculosis
by
Neena Sujata Kadaba
Submitted to the Biological Engineering Division on May 9, 2003
in Partial Fulfillment of the Requirements for the
Degree of Master of Science in Molecular and Systems Toxicology
Abstract
The purpose of this work is to probe the dTDP-L-rhamnose pathway in an effort
to develop small molecule inhibitors that could act as therapeutics for Mycobacterium
tuberculosis. The necessity for newer, more effective treatments for tuberculosis is
growing, as the bacteria evolve resistance to traditional treatments.
In an effort to develop more effective and perhaps more abbreviated courses of
treatment, a plan was developed to investigate a pathway involved in cell wall
biosynthesis as a promising target: the dTDP-L-rhamnose pathway. This pathway
plays an essential role in linking the peptidoglycan and arabinogalactan portions of the
mycolic acid-arabinogalactan-peptidoglycan complex, a significant part of the
mycobacterial cell wall. The mounting level of biochemical understanding of this
pathway and its importance in bacterial cell wall biosynthesis indicates that it is not
only a relevant target but also an accessible one. Of the four enzymes crucial to this
biosynthetic pathway, one was chosen as the primary focus: dTDP-D-glucose-4,6dehydratase (RmlB). There are 3 steps in the reaction mechanism of RmlB: oxidation
of the C4 position of dTDP-D-glucose to form a 4-keto structure, dehydration of the C6
position via the elimination of water and a subsequent reduction to result in a 6-deoxy
product. Crystal structures of this particular enzyme, dTDP-D-glucose 4,6-dehydratase
(RmlB), complexed with single substrates or substrate analogs have provided a
foundation for these studies, enabling the rational design of a small library of potential
inhibitors. Twelve mechanism-based inhibitors of RmlB are proposed. These
compounds reflect the current understanding of the mechanism and mimic the sugar
portion of the sugar-nucleotide substrate at various steps throughout the reaction
mechanism. Each of the proposed inhibitors is designed to inhibit one of the specific
steps of the mechanism. While the intention of this project is to synthesize each
compound in this library from commercially available starting materials in 15 steps or
less, the primary goal of this particular dissertation is to synthesize 3 of the 12 proposed
inhibitors from the commercially available starting material 1,5-anhydro-D-glucitol.
The long term goal of this work is to produce these compounds in significant amounts
in order to test their efficacy in an animal model of mycobacterial infection.
Thesis Supervisor: John M. Essigmann
Title: Professor of Chemistry and Toxicology
Acknowledgements
Above all, I would like to take the time to sincerely thank my advisor, John
Essigmann, for everything he has done for me over the past 4 years. I had the good
fortune of having John as my undergraduate advisor, and from that time, he has been a
consistent source of information and support. In the past year and a half in particular,
he has given me a number of opportunities that I greatly appreciate-it is thanks to him
that I had a wonderful time helping to teach the Biotechnology and Engineering course
in Bangkok, and the amazing experience of spending a few months in Thailand. It was
also John who had the great idea to attempt this research project this year, and I have
only become even more excited about the project as the year progressed; all of these
experiences have been a big part of my professional and personal growth over the past
year, and I am grateful for his influences.
I am eternally indebted to A. Nicole Dinaut, whose constant support and
encouragement over the past months has been invaluable. She has not only provided
guidance with every one of my chemistry dilemmas, but has also been a close friend,
always celebrating our successes and dismissing our setbacks with smiling optimism. I
am grateful to Kaushik Mitra for sharing his practical expertise, Denise Rodrigues for
her help with my syntheses and 2D NMRs, Uday Sharma for synthetic words of wisdom
and his assistance with hydrogenations and Will Neeley for sharing suggestions on
technical matters. Many other lab members were a pleasure to work with, and I thank
them for their moral support and friendship.
I would like to thank the students of BEH.214 during the spring of 2002 for
their assistance in putting the project together, and, in particular, Hector Hernandez for
working with me on the final project for the class. I would like to thank Drs. Mathuros
Ruchirawat and Somsak Ruchirawat for providing me the opportunity to stay and work
at the Chulabhorn Research Institute in Bangkok, Thailand and Kim Bond Schaefer for
continual administrative support while I was away. I appreciate the financial support
from the NIH Training Grant in Environmental Toxicology along with help from
Professor Douglas Lauffenburger of the Biological Engineering Division; both helped
make my trip to Thailand a reality.
Finally, I thank my family for their unconditional love and support. My mother
and grandparents have encouraged every one of my endeavors, often making personal
sacrifices for my benefit. They have instilled in me a love of learning, the confidence to
pursue my goals relentlessly, and the desire to do something meaningful.
Table of Contents
Design and Synthesis of Inhibitorsof dTDP-D-glucose 4,6-dehydratase (RmlB), an
Enzyme Required for dTDP-L-Rhamnose Production in M .Tuberculosis..........................
A b stract ....................................................................................................................................
Acknowledgements ...........................................................................................................
Table of Contents.....................................................................................................................
List of Fig u res...........................................................................................................................
List of Abbreviations .........................................................................................................
Chapter 1. Introduction...................................................................................................
1.1 Overview .....................................................................................................................
1.2 History and Epidemiology of TB Infection W orldwide .......................................
1.3 Infection with Mycobacterium Tuberculosis.......................................................
1.3.1 Characteristics of Mycobacterium Tuberculosis....................
1.3.2 The Mycobacterial Cell Wall .........................................................................
1.3.3 Transmission, Infection and Disease Progression .........................................
1.4 Host-Pathogen Interactions and the Immune Response ......................................
1.4.1 Macrophage Activation and Response ...........................................................
1.4.2 Factors increasing susceptibility.....................................................................24
1.5 Current Treatments and the BCG Vaccine...........................................................
1.5 .1 Ison iazid ...............................................................................................................
1
2
3
5
7
8
10
10
11
15
15
16
17
18
19
25
25
1.5.2 Rifampin .........................................................................................................
26
1.5.3 Ethambutol .......................................................................................................
1.5.4 Pyrazinamide...................................................................................................
1.5.5 Streptomycin ..................................................................................................
1.5.6 Second line drugs............................................................................................
1.5.7 Bacillus Calmette-Gudrin Vaccine ...............................................................
1.5.8 The Need for New Therapeutics ........................................................................
Chapter 2. Target Selection and Project Design .............................................................
2.1 Target Exploration: Narrowing Down the Possibilities.......................................
2.2 Target Exploration: Focus on Cell Wall Biosynthesis and the MAPc ..................
2.2.1 Peptidoglycan Biosynthesis ...........................................................................
26
27
27
27
28
28
30
30
33
33
2.2.2 Mycolic Acid Biosynthesis..............................................................................
34
2.2.3 Arabinogalactan Biosynthesis .........................................................................
2.3 Target Identification: The d-TDP-L-Rhamnose Pathway....................................
2 .4 Rm lB ............................................................................................................................
2.4.1 The Structure of RmlB ....................................................................................
2.4.2 The Active Site ................................................................................................
2.4.3 M echanism of Action..........................................................................................
2.5 Design of Inhibitors.................................................................................................
34
35
36
36
37
38
39
2.5.1 Inhibition of Oxidation ..................................................................................
2.5.2 Inhibition of Dehydration .............................................................................
41
41
2.5.3 Inhibition of Reduction...................................................................................
2.5.4 Selection of Experimental Focus ....................................................................
42
42
Chapter 3. Experimental M ethods...................................................................................
43
3 .1 Gen eral........................................................................................................................
3.2 Synthesis of 6-deoxy- 1,5-Anhydro-D-glucitol (10). .........................................
2,3-O-Acetyl-4,6-O -Benzylidene- 1,5-Anhydro-D-glucitol (19). ......................
43
43
43
2,3-O-Acetyl-4-O-Benzyl-1,5-Anhydro-D-glucitol (20) ....................................
44
2,3-O-Acetyl-4-O-Benzyl-6-Deoxy-6-Iodo- 1,5-Anhydro-D-glucitol (21)..........45
2,3-O-Acetyl-6-Deoxy~ 1,5-Anhydro-D-glucitol (22) .........................................
45
6-Deoxy- 1,5-Anhydro-D-glucitol (10). .................................................................
46
Chapter 4. Results and Discussion...................................................................................
47
4.1 Retrosynthetic Analysis .........................................................................................
47
4.2 Reduction of Methyl ax-D-Glucopyranoside .......................................................
49
4.3 Benzylidene Acetal Formation and Acetyl Protection.........................................
50
4.4 Assignment of NMR Spectra and the Regioselective Ring Opening Reaction.......51
4 .5 lod in ation ....................................................................................................................
55
4.6 Hydrogenation .......................................................................................................
56
4.7 Final Deprotection and Peak Assignment of Inhibitor 10 ..................................
57
4.10 Conclusions and Future Studies.........................................................................
58
R eferen ces ..............................................................................................................................
60
Fig u res....................................................................................................................................6
3
List of Figures
Figure 1. WHO statistics on global tuberculosis infection. ..........................................
64
Figure 2. Scanning electron micrographs of Mycobacterin tuberculosis..................... 65
Figure 3. Schematic representation of the cell wall of M. TB. .....................................
66
Figure 4. Stages of the immune response.......................................................................
67
Figure 5. The immune response against M. TB. ............................................................
68
Figure 6. Chemical structures of first-line drugs...........................................................
69
Figure 7. The glyoxylate pathway as it relates to the citric acid cycle.........................
70
Figure 8. Biosynthesis of the mycolic acid-arabinogalactan-peptidoglycan complex... 71
Figure 9. Biosynthesis of dTDP-L-rhamnose from glucose- I -phosphate.................... 72
Figure 10. The Structure of RmlB. ...................................................................................
73
Figure 11. The active site of Rm lB. .................................................................................
74
Figure 12. Proposed mechanism of action of RmlB. .....................................................
75
Figure 13. The twelve inhibitors designed for RmlB.......................................................
76
Figure 14. Inhibitors designed against the oxidation step of the mechanism ............. 77
Figure 15. Inhibitors designed against the dehydration step of the mechanism........ 78
Figure 16. Inhibitors designed against the reduction step of the mechanism............. 79
Figure 17. Retrosynthetic analysis using a protected glucal as the starting material.....80
Figure 18. Retrosynthetic analysis using D-glucose as the starting material..............81
Figure 19. Synthesis of inhibitors 10, 9 and 12. ............................................................
82
Figure 20. Suggested mechanism for the regioselective benzylidene ring opening
reactio n ..........................................................................................................................
83
Figure 21. 'H Spectrum of Compound 19, 2,3-O-acetyl-4,6-O-benzylidene- 1,5anhydro-D -glucitol. .................................................................................................
84
Figure 22. 13 C Spectrum of Compound 19....................................................................
85
Figure 23. gCOSY spectra of Compound 19. ................................................................
86
Figure 24. HETCOR spectrum of Compound 19...........................................................
87
Figure 25. IH spectrum of Compound 20, 2,3-O-acetyl-4-O-benzyl- 1,5-anhydro-Dglucito l...........................................................................................................................8
8
Figure
Figure
Figure
Figure
Figure
26. 13C spectrum of Compound 20. ...................................................................
89
27. gCOSY spectra of Compound 20. ................................................................
90
28. HETCOR spectrum of Compound 20...........................................................
91
29. 1H Peak assignments for compounds 19 and 20. .......................................
92
30. IH spectrum of Compound 21, 2,3-O-acetyl-4-0-benzyl-6-deoxy-6-iodo-
1,5-anhydro-D -glucitol ...........................................................................................
Figure 31. 13 C spectrum of Compound 21. ...................................................................
Figure 32. 'H spectrum of Compound 22, 2,3-O-acetyl-6-deoxy- 1,5-anhydro-D-
glucito l...........................................................................................................................9
Figure
Figure
Figure
Figure
Figure
33.
34.
35.
36.
37.
13 C spectrum of Compound 22. ...................................................................
93
94
5
96
1H spectrum of Compound 10, 6-deoxy- 1 ,5-anhydro-D-glucitol...........97
13C spectrum of Compound 10. ...................................................................
98
gCOSY spectrum of Compound 10..............................................................
99
HETCOR spectrum of Compound 10..............................................................100
List of Abbreviations
BCG bacillus Calmette-Gudrin
CTLs cytotoxic T-lymphocytes
DOTS Directly Observed Treatment, Short-course
dTTP deoxythymidine triphosphate
EMB ethambutol
FAS fatty acid synthase
ICL isocitrate lyase
IFN-y interferon gamma
IL- 12 interleukin 12
INH isoniazid
KO knock out
M. TB Mycobacterium Tuberculosis
MAPc mycolic acid-arabinogalactan-peptidoglycan complex
MDR-TB Multi-Drug Resistant Tuberculosis
MS malate synthase
NO nitric oxide
PI phosphatidyl inositol
POA pyrazinoic acid
PPD purified protein derivative
PZA pyrazinamide
RIF rifampin
RmlB dTDP-D-glucose 4,6-dehydratase
ROI reactive oxygen intermediates
RNI reactive nitrogen intermediates
SDR short-chain dehydrogenase/reductase
TACO tryptophan aspartate containing coat
TB Tuberculosis
TLRs Toll-Like Receptors
TNF-a Tumor Necrosis Factor alpha
UDP-Galf UDP-galactofuranose
UDP-Galp UDP-galactopyranose
UDP-MurNAc UDP-N-acetyl muramic acid
WHO World Health Organization
-9-
Chapter 1. Introduction
1. 1 Overview
Tuberculosis (TB), the disease caused by Mycobacterium tuberculosis (M.TB), is
currently a serious health problem in the developing world, while its threat in
developed countries continues to grow every year. Each year, over two million people
die from TB, including a large number of HIV-positive individuals who are unable to
fight the infection.1 Approximately two billion people are currently infected with M.
TB, a number that is constantly growing, as around nine million new cases are
identified each year. 2 This growing rate of infection translates into a new person
infected with TB every second; almost 1% of the total world population is newly
infected each year. Of the one-third of the world population infected with TB, 5-10%
will become sick or infectious during their lifetime.' The rates of infection are growing
at over 2% per year, despite implementation of health strategies worldwide.
Complicating matters is the sheer difficulty in treating the infection. Current
therapy consists of a six to nine month barrage of four or more different medications,
with different dosages and side effects. Despite this rigorous regimen, treatment can
still leave latent TB infection in the lungs. Adding to the current crisis is the growing
concern of multi-drug resistant (MDR) TB, which is defined as strains that are resistant
to both isoniazid and rifampin, the two most effective current TB drugs.
The World Health Organization (WHO), along with numerous health experts,
believe that the world is currently experiencing a public health crisis with respect to
TB, and that this problem will continue to persist unless measures are taken to better
control the spread of this disease. Unfortunately, 60% of TB-related deaths occur in the
poorest 20% of the global population, and, consequently, there is a serious dearth in the
amount of funding put towards profit-driven diagnostic and therapeutic development.
This project is derived from a novel approach to the development of therapeutics
for TB. Instead of focusing on an established project in an existing lab, the problem of
TB was analyzed in terms of identifying a new "ideal" target and developing a viable
method for altering its biological activity. To achieve this goal, the deficits in natural
and synthetic means to fight off the infection were examined. This was accomplished
through an analysis of both the mechanisms by which the mycobacteria are able to
overcome the human immune system as well as the shortcomings of existing
therapeutics. There are a large number of possible drug targets in mycobacteria.
While certain targets could be considered for further study, the main focus of this
project will be the selection of one particular enzyme as a target, and the remainder of
this dissertation will discuss the rational design of small molecule inhibitors of this
protein. In an effort to develop more effective and perhaps more abbreviated courses of
treatment, the target of interest is a particular enzyme, dTDP-D-glucose 4,6dehydratase (RmlB), involved in the dTDP-L-rhamnose pathway in cell wall
biosynthesis. Using the crystal structures of the enzyme complexed with single
substrates or substrate analogs as a foundation for the studies, the intentions of this
project were to conduct the rational design, synthesis and screening of a small library
of sugar analogs that could be potential therapeutics for TB.
1.2 HistoryandEpidemiologyof TB Infection Worldwide
The problem of TB is nothing new. TB is clearly a significant worldwide killer,
and has been for centuries; in fact, it has been referred to as the "white plague" due to
the pale appearance of its victims. Robert Koch, who discovered the TB bacilli, wrote in
1880, "If the number of victims which a disease claims is the measure of its
-1 1-
significance, then all diseases, particularly the most dreaded infectious disease, must
rank far behind tuberculosis. Statistics teach that one-seventh of all human beings die
of tuberculosis, and that, if one considersonly productive middle-age groups,
tuberculosis carried away one-third and often more of these..." As it was in the 19th
century and before, TB continues to be a problematic infection.
While cases of TB consistently dropped once therapeutics were introduced in the
1950's and 1960's, since the 1980's the number has been growing once again. There
are numerous possible reasons for this increase, but one of the main contributing
factors to an increase in TB infections worldwide is the HIV/AIDS epidemic, as those
afflicted with HIV/AIDS are much more susceptible to active TB infections than other
individuals. About one-third of those afflicted with HIV/AIDS are infected with TB.
Due to their compromised immune systems, HIV/AIDS victims are unable to fight the
infection, resulting in high rates of reactivation TB and high rates of fatalities (raised by
5-20% over the normal 15%) due to TB.3 Other possibilities for the increase in TB
infections worldwide could include increased poverty, higher levels of global travel,
and significantly higher levels of crowding in big cities, as well as institutions such as
prisons and hospitals. Lack of available treatment is another concern in the developing
world, as there is a 50% fatality rate for untreated cases-a number which rises to 80%
in HIV/AIDS patients. Whatever the reasons for this increase, it is clear that the
problem is only getting worse at the status quo. During the 1990's, it was
approximated that there were ninety million new cases and around thirty million
deaths due to TB. However, it is estimated that over the next twenty years,
approximately one billion more people will become infected, with more than a hundred
and fifty million people experiencing symptoms and over thirty six million of those
succumbing to the infection. The infection is currently the leading cause of death due
to infectious disease among people older than five. 4
Regionally, the distribution of TB is quite striking, as the rates of TB infection
worldwide vary tremendously. The prevalence of TB is much higher in developing
countries, as is the death rate due to the infection. TB accounts for 20% of adult deaths
and 6% of infant deaths, and is considered to be the cause of 26% of all avoidable
deaths in the developing world.' It is hypothesized that the developing world is home
to 95% of the world's TB, and 98%of the deaths caused by the disease; overall,
approximately 7% of all deaths in the developing world are attributed to TB. There is a
relationship between lower standards of living and higher rates of TB, as it has a much
higher incidence in poorer countries and large, crowded cities. Of the twenty two
countries with the highest occurrence of TB, only one (Brazil) was not classified as a
low income or low middle income country, according to the per capita income
designations set by WorldBank.6
The continents of Asia, Africa and South America are hardest hit, yet the top five
infected countries (India, China, Indonesia, Bangladesh and Nigeria) are almost all in
Asia (Figure 1). 1 About three million cases per year occur in south-east Asia, while
around two million cases per year are present in sub-Saharan Africa, an area of
increasing concern because of the growing relationship between HIV/AIDS and TB.
Forty percent of the global incidence occurs in India and China, the two countries with
the largest populations worldwide. India has an estimated 1.8 million cases while
China reports approximately 1.4 million. Both countries claim significantly high levels
of treatment success (79% and 96%, respectively), yet it is clear their National TB
programs are not as extensive and treatment is not as readily available as it should be.'
DOTS or Directly Observed Treatment, Short-course is the primary strategy
endorsed by the WHO to achieve high levels of treatment success. The strategy
encompasses several goals that involve government commitment in TB control
activities, case detection by sputum smear microscopy, standardized chemotherapy
~13-
regimens for eight weeks, an uninterrupted supply of essential anti-TB drugs, and
finally, a recording/reporting system allowing for assessment of each individual patient
and the TB program overall. It has been rated as an extremely cost-effective way of
efficiently treating large numbers of patients, yet the distribution of DOTS enforcement
is not as high as desired.! The countries listed above are not the countries with the
highest rates of incidence per 100,000 people. Cambodia, Zimbabwe, South Africa,
Afghanistan and Uganda are the countries with very high rates of disease and these
countries have very low therapeutic effectiveness. The enforcement of DOTS in most of
these countries is minimal or non-existent, due to funding and personnel shortages,
leading to the further spread of disease.'
Problems with effective TB therapy and DOTS enforcement have led to the
spread of multi-drug resistant (MDR) forms of TB. Due to the long course of treatment,
patient compliance is a serious issue, as patients tend to feel better after two months of
treatment. However, discontinuing treatment at that stage leads to the development of
strains of MDR-TB. Poor availability of therapeutics along with incorrect single-drug
treatment regimens also contribute to the growing problem of MDR-TB. There are
cases of MDR-TB on every continent, yet there are certain countries with particularly
high levels. Latvia (14.4%), Estonia (10.2%), Dominican Republic (6.6%), Ivory Coast
(5.3%), Argentina (4.4%) and Russia (4.0%) are among those with the most significant
incidences of MDR~TB. In countries not implementing DOTS, rates of MDR-TB greater
than 2% were much more common-54% in non-DOTS countries versus 22% in DOTS
countries. It is estimated that up to fifty million people may already be infected with
MDR strains of TB, and this number continues to rise. 7
-14-
1.3 Infection with Mycobacterium Tuberculosis
Infections due to M.TB pose a significant public health concern, yet it is the lack
of understanding of the complexities of the bacterium itself that frustrates the efforts to
contain the disease. The unique features of the pathogen pose as much of a challenge
in the development of therapeutics now as they have for the past 100 years. Despite
the vast knowledge base currently available regarding M. TB, there are a considerable
number of questions that remain. However, there have been recent developments in
understanding the interactions between the pathogen and the host that present a
beautiful scientific problem.
1.3.1 Characteristics of Mycobacterium Tuberculosis
Mycobacterium Tuberculosispossesses a variety of unusual characteristics that
contribute to its virulence as a pathogen. The bacterium has a Gram-positive type cell
wall, yet cannot be detected by the conventional Gram stain due to the waxy outer
coating of numerous lipids. The rod-shaped bacteria (ranging between 1 to 4 tm in
length and 0.3 to 0.6 pm in width) can be detected using an acid-fast stain, a harsh
stain that involves heating the bacteria. The primary method of diagnosis in most of the
world is through the culture of sputum samples and detection using this stain (Figure
2). The slow generation time of the bacterium (18-24 hours) is attributed to the
hydrophobic cell surface, which may cause the clumping together of bacteria, thereby
preventing the entry of nutrients into the cell. 2 The organism is an obligate aerobe,
and, as expected, it prefers the high oxygen content of organs such as the lungs, despite
having the capacity to change its metabolism and survive in a microaerophilic
environment.
-15-
1.3.2 The Mycobacterial Cell Wall
While M. TB is considered Gram-positive, the organism possesses cell-wall
features that are characteristic of both Gram-positive and Gram-negative bacteria,
resulting in a particularly unique structure. The mycobacterial cellular envelope
consists of three components: 1) a plasma membrane, 2) a covalently linked mycolic
acid-arabinogalactan-peptidoglycan complex (MAPc) and 3) a polysaccharide-rich
capsule-like material (Figure 3). The lipid content of the cellular envelope is so
dramatic that it is estimated to make up 30-40% of the total weight of the bacilli and
could account for characteristics such as low permeability, growth in clumps, physical
strength and a decreased response to certain toxic substances. 9
The plasma membrane serves as the initial protection of the mycobacterium
from the outside environment. It appears to have characteristics quite similar to
normal bacterial membranes, as does the peptidoglycan component of the MAPc. The
MAPc consists of the cross-linked peptidoglycan which is bound covalently to
arabinogalactan chains through a phosphoryl-N-acetylglucosaminosyl-rhamnosyl
linker. The arabinogalactan is additionally esterified to a variety of ca-alkyl and
p-
hydroxy mycolic acids. 10 In particular, the arabinogalactan-peptidoglycan component
of the cell wall has been implicated in the viability of mycobacteria; the relatively recent
elucidation of biosynthesis pathways has resulted in a number of targets for small
molecule enzyme inhibition. Certain pathways are becoming increasingly wellcharacterized, but the specific roles of many enzymes involved in these pathways
remain elusive. There are a number of lipids that are non-covalently attached to the
MAPc, and beyond this layer of lipids lies a loosely attached layer of polysaccharides
and proteins, forming the outer capsule around the mycobacteria. 9 Overall, this
extensive cell wall decoration plays a role in the behavior of mycobacteria in the host
environment, and could affect processes ranging from intracellular growth, to the
immune response, to reactionsof the bacilli to therapeutics.
1.3.3 Transmission, Infection and Disease Progression
The disease itself is spread primarily by aerosolized droplets of tubercle bacilli
containing three or less organisms; these droplet nuclei float in the air, and can remain
airborne for long periods of time after the fluid evaporates. A patient with pulmonary
TB can infect another while coughing, exhaling, sneezing, talking or spitting; it is
estimated that 10- 15 people each year are infected by one infected individual. 4
Once inhaled, the bacteria travel to the alveoli, as infection in the upper
respiratory tract is unlikely. Smaller droplets of bacilli can reach the alveoli, where
they are taken up by alveolar macrophages which attempt to phagocytose the bacteria.
Once the bacteria are taken up into the macrophages in a phagosome, the macrophage
attempts to lyse the bacteria by fusing the phagosome and the lysosome. This fusion is
often blocked by mycobacterial mechanisms, resulting in a pulmonary infection that
consists of active bacteria living within inactivated macrophages."I The mycobacteria
are able to grow and spread, and can even enter systemic venous circulation at times, so
they may re~enter the lungs and be deposited at additional locations within the lungs or
other organs. At this phase there are three possibilities for the mycobacteria and the
host effector cell: the macrophage can kill the mycobacteria, the mycobacteria can
cause the cell death of the macrophage, or the macrophage can become a host cell for
the mycobacteria and propagate infection."
It is at this point that an active immune
response is key in determining the outcome of infection. Poor T-cell immunity at this
phase can lead to a serious progression of the disease, as well as transmission to others.
The symptoms of an infection consist primarily of a fever, coughing for over three
weeks (often with bloody sputum), chest pain, loss of energy and weight loss, in
~17-
addition to progressive (irreversible) lung destruction. In most cases, the immune
response fails to completely clear the bacteria, which can continue to grow within
alveolar macrophages. 2
When phagocytes fail to clear the infection, other T-cells and macrophages
amass around the bacterial growth. Macrophages can fuse to form giant cells, resulting
in a layer consisting of macrophages and T-cells that surrounds the bacteria. This layer
often forms a self-contained lesion, consisting of a mass of bacteria surrounded by a
coat of fibrin. 2 This lesion, called a tubercle or granuloma, has a cheese-like
consistency (caseousnecrosis)at first, yet as bacterial growth continues and more
phagocytes surround the area, the lesion becomes more liquid. Latent TB consists of
mycobacteria contained in these lesions, and these liquefied lesions can form infectious
aerosols, burst to reactivate an active TB infection or travel to other organs, resulting in
miliary TB, which is often fatal. The chances of reactivation TB are 2-23% over a
lifetime, but this number can increase depending on immune system factors.2
1. 4 Host-Pathogen Interactions and the Immune Response
When encountering any pathogen, the human immune system attempts to
mount an effective response to kill the invader. The case of M. TB is no different; there
are a number of mechanisms by which the innate and adaptive immune systems can
clear the infection. However, what makes mycobacteria perhaps more complex than
certain other pathogens is their ability to evade damage by these intervention systems
and thus render immune system processes ineffective. This section will discuss some of
the interactions between mycobacteria and various human immune system components
in order to highlight the major players at the cellular and molecular levels.
1.4.1 Macrophage Activation and Response
Once phagocytosed bacteria are living within a macrophage, there are a
number of pathways that determine the fate of both the mycobacteria and macrophage.
The complexity of M. TB lies in its ability to modulate normal immunological activity in
its favor, despite a variety of mechanisms that are intended to clear the infection
through the various stages of infection (Figure 4).
Oxidative burst
Reactive oxygen intermediates (ROI) and reactive nitrogen intermediates (RNI)
comprise the two major sets of effector molecules important in the macrophage defense
against mycobacteria (Figure 5). These reactive intermediates are synthesized in
parallel by two related pathways. Beginning with 0z and L-arginine, the macrophage
is able to form radicals and other damaging agents to use against the bacteria. ROI are
a product descendent from 02, as the enzyme phagocyte oxidase (phox) catalyzes the
formation of superoxide, which is used in the synthesis of peroxide by superoxide
dismutase. Hydrogen peroxide then undergoes radical mechanisms to result in
hydroxyl and ferryl radicals. The importance of ROI in defense against mycobacteria is
slightly controversial, yet they do seem to play some kind of role, despite the number of
bacterial resistance genes and products that are used to neutralize this response. RNI
are formed in a similar manner from guanidine-L-arginine, which is transformed into
nitric oxide (NO) radicals due to nitric oxide synthase (iNOS or NOS2). These nitric
oxide radicals can go on to form peroxynitrite (along with superoxide) or they can
form N02O,
N02, S-nitrosothiols, dinitrosyl iron complexes and other reactive
molecules. Similar to ROI, mycobacteria have developed a number of mechanisms
involving molecular scavengers, antioxidant enzymes, repair systems and antioxidant
regulons."I Studies of inducible nitric oxide synthase (iNOS or NOS2) and phagocyte
NADPH oxidase (phox) suggest that RNI are significantly more important in the defense
against TB. Macrophages from actively infected individuals have been found to have
high levels of NOS2, and the levels of NO exhaled in TB patients are higher than that of
non-infected individuals. There appear to be more mechanisms by which ROI can be
evaded, as many cell wall components such as lipoarabinomannan and
phenolicglycolipid I are oxygen radical scavengers. However, despite the controversy
over the relative importance of these two mechanisms, it is clear that they play a role in
the macrophage defense.
Controlofphagolysosome fusion
The response of macrophagesto mycobacteriais to form a phagosome around
the bacteria, and once the bacteria have been taken up into the cell, fuse the
mycobacteria-containing phagosome with the acidic lysosome, thereby ensuring the
death of the bacteria. However, one of the most complex mechanisms of mycobacterial
survival is its ability to block this phagosome-lysosome fusion, allowing the bacilli to
survive indefinitely within the macrophage.
There are a number of biochemical reasons for the arrest in phagolysosome
maturation. The ability of the lysosome to degrade particles depends upon hydrolases
that must be maintained at low pH levels, yet it has been demonstrated that
mycobacteria-infected phagosomes possess a lack of vacuolar ATPases, which are
usually responsible for acidification; the less acidic environment could be responsible
for the lower amount of degradation by the vacuole.II Another possible mechanism is
the role of Rab GTPases; Rab5 helps in early endosome fusion while Rab7 assists in late
endosomal membrane trafficking, yet it has been found that mycobacterial phagosomes
utilize Rab5 while they exclude Rab 7.13 Calcium-binding calmodulin may be involved
in the regulatory functions of membrane fusion, as phagocytosis usually leads to an
~20~
increase in calcium concentrations, yet mycobacteria are able to alter the signaling
cascade and prevent this increase.'I Another important protein could be the
tryptophan aspartate containing coat (TACO), an actin-binding protein that is
implicated in the early phagosome membrane. However, the TACO-membrane
association is expected to be transient, yet the retention of this association is postulated
to prevent phagolysosome fusion by delaying maturation of the phagosome;
mycobacteria that are somehow able to prevent the detachment of TACO are able to
survive longer within macrophages.12 There are other mechanisms that may explain
the lack of degradation of mycobacteria by macrophages; certain mycobacterial
sulfatides may be able to inhibit phagolysosomal fusion, while it is also possible that
high levels of ammonia production by mycobacteria may change saltatory movements
and even raise the pH of the compartment. 1 3
CD8+ Cells
The activities of CD8+ cells, in protecting against mycobacteria are two-fold; the
cytotoxic functions of the cells enable them to kill intracellular mycobacteria directly in
addition to serving as a facilitator in lysing macrophages containing live bacteria
(Figure 5). Initially, the CD8+ cells are presented with mycobacterial lipid or peptide
antigens by antigen-presenting cells via molecules such as MHC Class I or CD 1.14
Once these CD8+ cells come into contact with an infected macrophage, they can
activate the macrophage to kill the mycobacteria through the production of interferongamma and tumor necrosis factor-alpha, two crucial cytokines that can activate
enzymes such as nitric oxide synthase. CD8+ cells are able to function as cytotoxic T
lymphocytes (CTLs) in order to facilitate lysing of infected macrophages. This lysing
can occur via two pathways: a perforin-dependent pathway and a Fas/FasL pathway.
The perforin dependent pathway involves granulysin: perforin forms a pore that
-21~
enables the uptake of granulysin, a cytolytic T cell granule protein, into the
macrophage, resulting in the death of both the macrophage and its mycobacteria.
However, if the mechanism of macrophage lysis is via the Fas/Fas ligand system, the
macrophage is destroyed but the live mycobacteria are released.14
CD4+
CD4+cells are important in the presentation of mycobacterial antigens by MHC
class II molecules. However, the primary effector function of CD4+ cells is the active
role in IFN-y production, which is crucial in macrophage activation. These cells are
also involved with NOS2 expression in the early phases of infection, but these cells play
a role in later stages of the infection as well, as they have been found to be important in
the priming and maintenance of CD8+ T cell effector and memory functions. The
influence of these cells in the immune response has been studied extensively. For
example, it has been found that M. TB infected mice treated with an anti-CD4+
antibody experienced a reactivation of active disease, demonstrating the key role of
these cells. M. TB has been able to alter the activity of these cells through disrupting
the macrophage-CD4+ cell interaction. M. TB-infected macrophages are unable to
present antigens to CD4+ T cells, and infection by mycobacteria may inhibit recognition
of macrophages by CD4+ cells by down regulating the surface expression of MHC Class
II molecules. This interference with normal function is yet one more mechanism by
13
which mycobacteria can survive within the host successfully.
Interferon-gamma
Interferon-gamma (IFN-y) is perhaps the most important cytokine involved in
M.TB infection. It plays a crucial role in macrophage activation by enhancing the Th 1
response and increasing IL- 12 production.' 3 It is produced by CD4+ and CD8+ cells,
along with NK cells, and has been found to be important in antigen-specific T-cell
immunity.13 The functions of IFN-y are quite diverse; it has been known to promote the
production of opsonising antibodies, increase CTL activation as well as NK cell activity
and increase macrophage MHC Class II activity.' 6 It may also trigger the movement of
T cells into infected areas by facilitating a bond between endothelial cells and T cells
7
and serve as a trigger to differentiate macrophages into active effector cells.1 It has
been found that IFN-y knock-out (KO) mice are most susceptible to virulent TB;
additionally, it was found that KO mice had a defective macrophage response and low
NOS2 expression.13
Tumor Necrosis Factor-a
Tumor Necrosis Factor-ca (TFN-a) is a cytokine that plays a role in the mediation
of macrophage activation.'
3
During acute infection of M. TB, TNF-ca expression is
increased. Along with IFN-y, TNF-a initiates an increase in NOS2 expression, enabling
a stronger response to M. TB infection. TNF-a may also coordinate the migration and
localization of cells within particular tissues, and it appears to be essential in
granuloma formation, as it regulates expression of certain adhesion molecules.
However, it has also been found that TNF-cL is responsible for lung tissue damage, as
well as the symptoms of weight loss experienced by TB patients.
Interleukin- 12
Interleukin- 12 (IL- 12) is a pro-inflammatory regulatory cytokine which is
primarily produced by phagocytes.I 7 The release of IL- 12 is induced after
phagocytosis, and is thought to initiate a Th 1 response by stimulating IFN-y
production.' 3 ThLC
wCAct role 0fpr
oT
1
i
not
completely
IL- Iz Is 1 ot compieLtIy
-23-
u
s
4IiLUL;-LVULL
yet
Y%,L
i
ItL is
,I
17
found that IL~ 12 KO mice are highly susceptible to M.TB infections, while treating TB
infected mice with IL- 12 in the early phase of infection increased the mean survival
time and decreased the bacterial load.13 IL- 12 serves as a mediator that can connect
the response of phagocytes to that of T cells through its control over IFN~y.
Toll-Like Receptors
Adaptive Toll-like receptors (TLRs) facilitate the activation of the immune
response against M.TB. TLRs can recognize a number of mycobacterial antigens; the
antigens most important in the immune response are the 19kDa lipoprotein,
lipoarabinomannan and phosphatidylinositolmannan, all of which interact with TLR2.
TLR2 is proposed to have a more significant role in defense against TB than TLR4;
activation of TLR2 can result in an induction of macrophage apoptosis and the down
regulation of MHC class II, two important anti-mycobacterial functions.18
1.4.2 Factors increasing susceptibility
Most characteristics that have been identified as increasing a particular
individual's susceptibility towards TB infection involve the response of the immune
system; whether it is in the form of a genetic variation or other factors that decrease
normal immune functions. A defect in IL- 12 or IFN-y causes an individual to be at a
significantly heightened risk of TB, while other mutations, such as ones in the promoter
regions of TNF-c, or the natural resistance-associated macrophage protein 1 (nRAMP),
result in only slight increases in susceptibility.19
A normal person infected with TB bacilli incurs a 10% risk of developing active
TB throughout a lifetime. A weakening of the immune system, however, can increase
this risk. The single largest factor to increasethe chances of reactivation is HIV
infection; the risk level is raised to 8-10% per year (as opposed to 10% per lifetime) due
4
to the immunocompromised state of the patients. Other factors include silicosis, renal
disease, malignancy, diabetes, malnutrition or long-term treatment with
immunosuppressive drugs.19
1.5 Current Treatments and the BCG Vaccine
From the above description of mycobacteria, it is quite clear that there are a
number of complexities associated with successfully fighting infections of M. TB. There
are a number of anti-mycobacterial agents in use, yet none are perfect. In fact, only
four can be considered "first-line" treatments, while others are only used on an "asneeded" basis due to the rapid resistance that is developed by the mycobacteria. TB
therapy is currently made up of an initial two-month phase of four drugs daily
(Isoniazid, Rifampin, Pyrazinamide and Ethambutol) after which time the regimen
solely consists of Isoniazid and Rifampin. In cases of MDR-TB, the so-called "DOTSplus" strategy is employed, and the use of second-line drugs along with any viable firstline options is used. This section will summarize the currently available treatments and
the problems associated with each.
1.5.1 Isoniazid
Isoniazid (INH) has been the most important anti-mycobacterial agent in use
since the 1950's, and it continues to dominate therapeutic regimens, aside from cases of
MDR-TB (Figure 6). It is hypothesized that INH is a prodrug which is transformed into
a number of reactive radical species by the mycobacterial catalase-peroxidase KatG.
These reactive molecules can react with a number of mycobacterial targets, but the two
most well-understood enzymes are InhA (an enoyl ACP reductase) and KasA (a beta-
ketoacyl ACP synthase), both of which are involved in the cell wall mycolic acid
biosynthesis. Resistance often arises to INH. Resistance to INH can be attributed to
mutations in katG; it is proposed that a particular serine is mutated into a threonine,
resulting in a form of the catalase-peroxidase that does not react with INH but
maintains enough normal activity to continue detoxification against antibacterial
radicals.2 0 Another gene, ndh (NADH dehydrogenase), which increases the ratio of
NADH/NAD+ in order to prevent a reaction between the isonicotinic acyl radical and
21
Despite the fact that the
an NAD radical, has also been implicated in INH resistance.
activated INH can react with a number of mycobacterial cellular components, drug
resistance is becoming a serious problem.zi
1.5.2 Rifampin
Rifampin (RIF), a compound with a very complex macrocyclic ring structure,
works to inhibit transcription in mycobacteria via an interaction with the DNAdependent RNA polymerase (Figure 6). The mode of action is fairly well understood,
and crystallographic studies confirm the tight binding of the drug to RNA polymerase.
The means by which mycobacteria develop resistance to rifampin is based upon
mutations in an 81 -base pair region of the gene encoding the B-subunit of RNA
polymerase (rpoB).20
1.5.3 Ethambutol
Ethambutol (EMB), a diamine derivative, is another first-line therapeutic used in
defense against TB (Figure 6). The precise mechanism of action is not yet known, yet it
is clear it is involved in the inhibition of cell wall biosynthesis. It has been shown to
inhibit the incorporationof mycolic acid into the mycobacterial cell wall, 2 and it has
more recently been suggested that EMB blocks some part of the polymerization process
involved in arabinogalactan and lipoarabinomannan biosynthesis. 23 More support for
an interaction between EMB and arabinosyl transferases is demonstrated by the gene
which confers resistance to EMB; mutations in embCAB, an operon encoding arabinosyl
24
transferases, have been implicated in most cases of resistance to EMB.
1.5.4 Pyrazinamide
Pyrazinamide (PZA) is a nicotinamide-like prodrug that reacts with
pyrazinamidase to form pyrazinoic acid (POA), the active form of the drug, that can
accumulate in the mycobacterial cytoplasm (Figure 6). Large amounts of POA can
cause a lowering of the intracellular pH, and perhaps inactivates pH sensitive enzymes
such as fatty acid synthase. When resistance to PZA occurs, it is usually the case that
mycobacteria stop using pyrazinamidase. Most PZA-resistant mycobacteria possess a
mutation in the promoter region of the gene that encodes pyrazinamidase (pncA).20
1.5.5 Streptomycin
Streptomycin, an aminocyclitol glycoside antibiotic, is considered as another
first-line drug against TB, yet its use is limited by serious toxicity and its intramuscular
delivery method (Figure 6). It inhibits protein synthesis by somehow interfering with
translational proofreading, perhaps by interacting with the ribosomal S12 protein and
16SrRNA. 2 1 Streptomycin increases the missense rate by suppressing proofreading and
possibly disrupts the decoding of aminoacyl-tRNA, leading to incorrect translation.
Resistance is attributed to point mutations in the S12 ribosomal protein encoded by
25
RpsL, and mutations in the rrs operon encoding the 16S rRNA.
1.5.6 Second line drugs
There are a number of second-line drugs that are used in the treatment of TB.
These therapeutics are classified as second-line because of decreased efficacy against
the bacilli, increased toxicity, increased rates of resistance or unfavorable
pharmacokinetics. These drugs include ethionamide, kanamycin, cycloserine, para-
-27-
aminosalicylic acid, capreomycin, amikacin and the fluoroquinolones, all of which are
used primarily in cases where resistance is developed against the first-line therapeutics.
However, none of these are optimal therapeutic options for the treatment of TB
infections. 2
1.5.7 Bacillus Calmette-Gudrin Vaccine
The vaccine available for TB, called bacillus Calmette-Gudrin (BCG), is an
attenuated strain of Mycobacterium bovis that is the most widely administered vaccine
in the WHO Expanded Programme for Immunization. However, the efficacy of this
vaccine is quite controversial. Protection from disseminated and meningeal TB in
infants and children is quite reliable, but its value in preventing pulmonary TB in adults
is questionable; studies on the rates of vaccine protection show dramatically different
8
efficiencies, ranging from 77%in the UK to 0% in India. As it stands, the BCG vaccine
does not do a successful job of providing immunity against TB, and is currently
contraindicated in HIV patients, as it is a live vaccine. 26 Additionally, vaccination with
BCG results in a positive response to the purified protein derivative (PPD), and
therefore a false-positive on the Mantoux diagnostic skin test, one of the most effective
diagnostic tools used to identify TB-infected persons.2
1.5.8 The Need for New Therapeutics
In 1993, the WHO declared that TB was a global emergency, yet over the past
ten years there has not been a significant increase in practical solutions to this problem.
There has not been a new drug approved for TB in over thirty years, yet the current
health crisis demonstrates a significant need for new developments in order to eradicate
TB. The effective therapies currently in use are not optimal; due to nine-month long
regimens, patient non-compliance is a serious issue, leading to the emergence of newer
drug-resistant strains. New anti-TB therapies are an absolute necessity for MDR-TB as
well as latent TB infection, and the development of such novel therapies is the focus of
the remainder of this thesis.
-29-
Chapter 2. Target Selection and Project Design
2.1 TargetExploration:NarrowingDown the Possibilities
In the process of choosing the dTDP-L-rhamnose pathway in TB as the
therapeutic objective for rational drug design, a range of possibilities were considered
based upon the current literature. The field of TB research is constantly evolving, and
the validity of targets is constantly changing. The purpose of this section is to briefly
explore various other possible targets and to explain reasons why the dTDP-L~
rhamnose pathway was viewed most favorably as a target for novel TB therapeutics.
IsocitrateLyase andMalate Synthase
The glyoxylate cycle is a shunt that is used to effectively convert acetate
(degraded from fatty acids, amino acids or ethanol) into carbohydrates by bypassing the
loss of CO 2 in the citric acid cycle in M. TB (Figure 7). The two enzymes involved in
the cycle, isocitrate lyase (ICL) and malate synthase (MS) catalyze the formation of
malate from isocitrate, via the intermediate glyoxylate; the cycle facilitates the
formation of succinate from two acetate molecules (as acetyl-Coenzyme A), conserving
carbon that would otherwise be released as carbon dioxide via the normal Krebs cycle
enzymes. In conditions of latency, fatty acids may be the primary source of carbon and
energy for mycobacteria, and it has been found that the ability of M. TB to persist in
27
macrophages is dependent upon ICL but not necessarily MS.
The use of ICL as a possible target for anti-mycobacterial development bears
further investigation, as not only is it an enzyme not present in humans, but it has been
examined extensively through mechanistic and crystallographic studies. However, it
was believed to an imperfect target because of the similarities between the substrate
and end products of the enzyme as compared to endogenous human substrates from the
citric acid cycle. Any inhibitors would most likely lead to undesired side effects.
Galactofuranosepathway
Another particular pathway of interest was the pyranose pathway consisting of
UDP-galactopyranose (UDP~Galp) mutase and galactofuranose (Galf) transferase,
which together function to transform an UDP-galactopyranose into UDPgalactofuranose and transfer galactofuranose molecules onto the growing chain of the
arabino-galactan complex, a component of the cell wall. The UDP-Galp mutase has
been crystallized, although the mechanism by which the mutase achieves the
unprecedented ring contraction of a nonreducing sugar is unclear. Galf transferase is
even more complex; it has been suggested that there are two to four separate
transferases involved, with the overall activity being significantly complicated. 28
Nevertheless, the pathway of core galactofuran synthesis has been found to be essential
for the viability of mycobacteria, and these galactofuranose units are not present in
human tissues. 29 Despite these important details, the lack of a clear mechanism renders
30
these enzymes a less attractive target for the rational design of inhibitors.
Antgen 85 complex
The Antigen 85 complex is made up of three related mycolyl transferases,
Antigen A, B and C. These three proteins are important in transferring a mycolic acid
molecule between one trehalose monomycolate molecule and another,3 ' which leads to
the formation of trehalose dimycolate, also known as the "cord factor," and free
trehalose, two important factors that are incorporated into the mycobacterial cell
wall. 3 2 The cord factor has been implicated in cell wall integrity, while the Antigen 85
complex itself may also foster interaction between the mycobacteria and the
macrophage.-" However, while there are some structures for this complex, the role of
these proteins is not completely understood; the value of this complex as a drug target
has been controversial, and thus it was decided to pursue alternatives.
Phosphatidylinositolsynthase
Phosphatidylinositol (PI), along with other lipoglycan molecules such as PI
mannosides, lipomannan and lipoarabinomannan, have been implicated in the cell wall
integrity of M. TB, and may additionally be important in the response of the
mycobacterium to the host immune system. Studies of PI synthase mutants have
revealed that this enzyme is essential to mycobacterial survival and the mechanism is
somewhat well understood, suggesting that it could be a viable therapeutic target.
However, as humans also possess a version of PI synthase, inhibitors of this enzyme may
33
lack specificity and could be quite harmful to human tissues.
Other Targets
The sheer number of possibilities for TB drug development is quite
overwhelming. Due to the recent sequencing of the M.TB genome, there have been a
number of targets identified, yet the mechanistic and structural knowledge about many
of these is still in the early phases of discovery. Among targets that could be important
are: the gated mechanosensitive ion channel, which has been implicated in ionic and
osmotic regulation; InhA, the enoyl-acyl carrier protein reductase; glutamine synthase,
which helps to control ammonia levels and thereby modulate phagosome-lysosome
fusion; and the
p-ketoacyl-ACP synthase, which is important in mycolic acid
biosynthesis. Numerous other pathways such as those involving peptidoglycan
synthesis specific to mycobacteria, and the
p-oxidation processes important in fatty acid
degradation could also reveal worthy targets. Another strategy employed to treat TB
involves the modulation of the host-pathogen interaction, in order to augment the
immune response, enabling the body to successfully fight the infection. It is clear,
however, that the knowledge base for this type of drug development is currently
lacking; not enough is known about such interactions to design appropriate
therapeutics effectively. Since further development is required before these strategies
can be successfully pursued, a more well-established pathway became the focus of this
project.
2.2 Target Exploration:Focus on Cell Wall Biosynthesisand the MAPc
As previously mentioned, the composition of the cellular envelope consists of
three components: 1) a plasma membrane, 2) a covalently linked mycolic acidarabinogalactan-peptidoglycan complex (MAPc) and 3) a polysaccharide-rich capsule
like material. The MAPc in particular is assembled via a complex set of biosynthetic
pathways; the three components are synthesized in parallel and joined together as the
final step.
2.2.1 Peptidoglycan Biosynthesis
Since peptidoglycan structures are fairly homologous between various bacterial
species, it is assumed that the peptidoglycan layer of M. TB is no different. Two
enzymes, MurA and MurB, are important in the synthesis of UDP-N-acetyl muramic
acid (UDP-MurNAc) from UDP-N-acetyl-glucose and enoylpyruvate. The UDPMurNAc is subsequently used to form a UDP-MurNAc-pentapeptide through the
linkage of an L-alanine, a D-glutamic acid residue, a diaminopimelic acid molecule and
finally a D-alanyl-D-alanine dipeptide, reactions catalyzed by MurC, MurD, MurE and
MurF. At this point, this peptide molecule is linked to an undecaprenyl phosphate and
bound to a N-acetyl glucose molecule to form a GlcNAc-MurNAc(pentapeptide)-
-33-
diphosphoryl-undecaprenol, also known as Lipid II, by MurG. This fragment is
subsequently ready for transport to the inner cell membrane where the undecaprenyl
diphosphate is released and the peptide crosslinks are formed. Each of these enzymes
are potential targets for anti-bacterial agents, yet one disadvantage is that there would
be no specificity for mycobacteria in particular.10
2.2.2 Mycolic Acid Biosynthesis
The mycolic acids decorating the surface of mycobacteria can be classified in a
variety of different ways, yet most possess a linear carbon chain (between 60 and 90
carbons), with vicinal carboxyl and hydroxyl groups on carbons 23 and 24 (±2), and
often contain cyclopropane rings in place of double bonds. 34 Both fatty acid synthase
(FAS) type I and type II systems are present in mycobacteria, and the synthesis of these
mycolic acids proceeds like much of fatty acid synthesis; the fatty acid synthase system
processes-p-ketoacyl synthase,
p-ketoacyl reductase, p-hydroxyacyl dehydratase and
enoyl reductase-perform the cyclic functions of condensation, ketone reduction,
dehydration and enoyl reduction. However, while the system is relatively wellcharacterized, not only are structural details not fully known, but there is additionally a
lack of specificity with regards to other bacteria (FAS II) or even humans (FAS I).
2.2.3 Arabinogalactan Biosynthesis
The portion of the MAPc that was of most interest was the branched-chain
arabinogalactan component, which links together the peptidoglycan and mycolic acid
layers. Two residues, a N-acetyl-glucose and a rhamnose, link the peptidoglycan and
arabinogalactan layers together; the galactose moieties are added on to this rhamnose
residue in succession, followed by the branching out of the layer by the addition of the
arabinose groups (Figure 8). The synthesis of the galactose molecules has been
~34-
described earlier, and it is hypothesized that the arabinose sugars originate from the
pentose phosphate pathway/hexose monophosphate shunt. While the synthesis of this
layer has been somewhat characterized, the majority of the enzymes involved have not
been crystallized or even distinctly isolated; it was decided that the optimal target for
this project would be a structurally and mechanistically understood pathway, and for
this reason many of the arabinogalactan component enzymes were eliminated.' 0
2.3 TargetIdentification:The d-TDP-L-Rhamnose Pathway
The arabinogalactan component of the cell wall is linked to the peptidoglycan
part via an N-acetylglucosamine and a rhamnose. As rhamnose is not a sugar
produced by humans, it is synthesized within the bacteria and transferred onto the
growing chain. L-Rhamnose is a 6-deoxyhexose that is synthesized from glucose-Iphosphate and deoxythymidine triphosphate (dTTP) and found in the cell walls of
numerous pathogenic bacteria. This pathway represents an interesting therapeutic
target, as no analogs of this pathway have been elucidated in humans. 35
The first enzyme in the pathway is RmlA (glucose- 1-phosphate
thymidylyltransferase), which transfers a dTDP on to glucose-I-phosphate to make
dTDP-glucose. RmlB, a dTDP-D-glucose 4,6-dehydratase, next converts the dTDPglucose into dTDP-4-keto-6-deoxy-D-glucose. The third enzyme in the pathway, RmlC
or dTDP-4 -keto-6-deoxy-D-glucose, epimerizes the dTDP-4-keto-6-deoxy-D-glucose
into dTDP-L-lyxo-6-deoxy-4-hexulose. Finally, RmlD catalyses the reduction of the C4
keto group to form dTDP-L-rhamnose (Figure 9).3 The dTDP-L-rhamnose is then
transferred onto the N-acetylglucosamine that will become the arabinogalactan
component. As an essential linker between the peptidoglycan and the arabinogalactan
molecule, rhamnose serves an important purpose; the inhibition of this sugar pathway
would theoretically prevent the assembly of the complex and prevent bacterial
survival. 10 It has been recently found that the formation of dTDP-L-rhamnose is
37
essential for the growth of M.TB.
2.4 RmIB
RmlB, the second enzyme in the dTDP-L-Rhamnose pathway, is perhaps the best
characterized in terms of structure and mechanism of the four enzymes. Of the four, it
presents the best therapeutic target because of the high level of understanding of its
function. Distinct crystal structures of the enzyme with its cofactor and the substrate
have been solved, resulting in considerable insight into the positioning of these
molecules within the active site.
Mechanistic studies have confirmed the presence of proposed intermediates, and
the convergence of both mechanistic and structural analyses have resulted in general
acceptance of the currently proposed mechanism. Additionally, the intermediates and
product of this conversion are distinct from any human substrates, suggesting that
selective targeting of the bacterial enzyme is possible. Therefbre, structure-based
inhibitordesign will be used to produce small-molecule inhibitorsthatare hoped to
selectively act upon this target enzyme.
2.4.1 The Structure of RmlB
RmlB is a member of the short-chain dehydrogenase/reductase extended family.
It is a homodimeric enzyme; monomer association occurs through hydrophobic
interactions via a four-helix bundle. Each monomer has two domains: a larger Nterminal Rossmann fold with seven
p-strands and ten a-helices and a smaller C-
terminal domain with six a-helices and four
p-strands. These domains come together
to form an active site cavity that encompasses both the sugar-nucleotide and cofactor
binding sites. The larger domain binds the cofactor NAD+ and the smaller domain
binds dTDP-D-glucose. The binding site contains a three glycine motif present in most
enzymes of the short-chain dehydrogenase/reductase (SDR) family and the SDRcharacteristic active site residues of threonine, tyrosine and lysine form a catalytic triad
that is essential for activity. The enzyme utilizes NAD+ but regenerates it at the end of
the reaction. The structure of RmlB from S entericawith the cofactor NAD+ and the
substrate bound has been reported at 2.57A (Figure 10).35
2.4.2 The Active Site
Within the active site, the NAD+ adopts a syn conformation; the si face is exposed,
leading to pro-S hydride transfer from the C4 sugar-nucleotide (Figure 11). There are
both polar and hydrophobic residues important in binding the NAD+. The NAD+
diphosphate group is dependent upon the characteristic SDR GlyXGlyXXGly sequence
for binding, while the adenine sits in a hydrophobic pocket around Ile21, AlaS7, Val77,
Ala81 and Leu 107. The nicotinamide ring shares electrostatic interactions with Asp32,
Thr35, Asp58, Thr99, Tyr167 and Lys 171, positioning it in the proper way for the
35
reaction mechanism.
Binding of the nucleotide-sugar is similarly stabilized by both hydrophobic and
polar interactions with nearby residues. The thiamine ring is surrounded by a
hydrophobic crevice created by Pro204, Leu207, Leu2 10, Val2l 1, Pro222, Ile223,
Val269 and Val270. Additionally the ring n-stacks with Tyr224. The sugar portion of
the substrate is involved with Thr133, Asp134 and Glu 135, each of which play
important roles in the reaction mechanism. Thr 133, along with Lys 171, help to lower
the pKa of Tyr 167, which is situated next to the C4 hydroxyl group. Thr 133 and
Glu 135 form hydrogen bonds with C4 while Asp 134 interacts with the hydroxyl group
located on C6. The sugar ring is on a parallel plane to the NAD+ nicotinamide ring;
this ring also serves to lower the pKa of the active site base Tyr 167.38
2.4.3 Mechanism of Action
This enzyme catalyses the conversion of dTDP-D-glucose to dTDP-4-keto-6deoxy-D-glucose through three primary steps: oxidation, dehydration and reduction
(Figure 12).
Oxidation
The first step in the mechanism is the oxidation of the glucosyl C4 and the
corresponding reduction of NAD+ to NADH. The process begins with a lowering of the
pKa of Tyr 167 to 6.41 due to hydrogen bonding to nearby Thr 133 and the positive
electrostatic fields created by nearby Lys206 and NAD+, allowing Tyr 167 to act as the
active site base. Once the substrate binds, the nicotinamide C4 is 3.5A away from the
glucosyl C4, positioning it for a hydride transfer. The tyrosinate deprotonates the
hydroxyl group on C4, leading to the formation of the ketone and reduction of the
NAD+ (compound 14 in Figure 12). This leads to a lowering of the pKa of the proton on
38
C5, which is important in the dehydration.
Dehydration
The dehydration step involves the removal of water across C5 and C6 and can
proceed via two possible mechanisms: the enolization of the dTDP-4-ketoglucose
intermediate, followed by elimination, or by a concerted elimination across the C5-C6
bond (compound 15 and 16 in Figure 12). It is not clear which mechanism is in effect,
but recent evidence suggests the latter may be occurring. The key in this step is to
protonate the leaving C6-hydroxyl group and deprotonate C5, allowing the formation
of the double bond. It is believed that Asp 134 donates a proton to the hydroxyl group
while Glu 135 abstracts the proton from C5.38
Reduction
In the last step, a hydride is transferred from NADH to C6, causing a
regeneration of the original NAD+ molecule. The end result is an inversion in the
configuration at C6, along with the addition of a proton back to C5 to make the final
product, dTDP-4-keto-6-deoxy~D -glucose (Compound 17 in Figure 12). Studies on the
distances between the C6 and the NADH nicotinamide ring suggest that they must be
close together, implying that the sugar changes position after the elimination in order
to be aligned properly for the reduction. The theory that the hexopyranosyl ring rotates
slightly, while the dTDP portion of the substrate remains in place, is supported by
experimental evidence. This movement would bring Tyr 167 close enough (3.5A) to
C5, indicating that it would be possible for Tyr 167 to protonate C5 and renew itself as
the active site base. It is also hypothesized that Glu 135 is within 3.5A of Asp134, and
38
the proton on Glu 135 could shuttle back to Asp 134 ending the catalytic cycle.
2.5 Design of Inhibitors
The library of compounds designed as proposed inhibitors of RmlB is derived
from the current understanding of the mechanism (Figure 13). Each of the molecules
proposed are sugar analogs; they possess no hydroxyl (or nucleotide group) at C1. The
1 -deoxy form is preferred as the lack of nucleotide results in a smaller molecule
overall, which is desirable in terms of cellular uptake and even drug delivery. In the
event that potent compounds are identified from this library, further structure-activity
studies will be proposed. The work discussed herein is therefore a starting point for the
design of effective and therapeutically viable inhibitors of RmlB.
As each step of the mechanism is targeted, the question of whether changes and
shifts within the enzyme active site during the reaction will affect the efficacy of these
molecules. While there is a proposed rotationof the bond between the sugar moiety
and the nucleotide portion of the substrate during the course of the reaction, this
supposed movement of the sugar-nucleotide during catalysis should not be an issue, as
it is a change in the phosphate bond between the sugar and nucleotide and these
proposed inhibitors do not have a nucleotide component. Therefore, it is hoped that
these molecules will be easily accommodated by the active site of the enzyme regardless
of which stage of the mechanism they are designed to mimic.
Liu et al have synthesized an inhibitor, CDP-6-deoxy-6,6-difluoro-D~glucose
against a related enzyme, CDP-D-glucose 4,6-dehydratase (isolated from Yersinia
pseudotuberculosis),that has activity of 0.94 mM (Ki) against the enzyme and forms a
covalent bond to the enzyme.-"9 This enzyme has an analogous mode of action to Rm1B,
and such a mechanism is purported to be similar in the formation of all 6deoxyhexoses. However, it is not clear how closely related the active site of that
enzyme is to RmlB, and therefore unclear as to whether these sorts of inhibitors would
have similar activity as inactivators of RmlB. Additionally, as aforementioned, the
analogs proposed do not have a nucleotide attached, as the nucleotide binding site is
removed from the active site residues. The inclusion of a nucleotide group would
increase the molecular weight of the inhibitor, which could lead to problems in cellular
uptake and bioavailability, therefore initial inhibitor design will focus solely on
modifications of the sugar moiety.
It remains unclear if these molecules will be taken up into mammalian cells, as
glucose uptake into cells depends upon the specificity of the glucose transporter and
molecules of glucose permease that allow passive diffusion into cells. However, since
certain modified glucose molecules (such as deoxy and fluorine substituted analogs) are
taken into cells, there is a high possibility that they can be taken up into macrophages
and even through the GI tract (if administered orally, for example). Bacterial cells
possess even more sugar transporters and permeases than mammalian cells, and are
therefore even more likely to take up these molecules. They depend on exogenous
sugar molecules for survival, and these molecules closely resemble the proposed
inhibitors, therefore it is possible that due to the carbohydrate-like nature of these
inhibitors, mycobacterial uptake may not be a serious issue.
2.5.1 Inhibition of Oxidation
Compound 1 replaces the proton of C4 with a fluorine, which could prevent the
initial oxidation in the first step of the mechanism (compound 14 in Figure 14) but may
still resemble the substrate closely enough to bind well to the active site of RmlB.
Compound 12 possesses an exocyclic carbon-carbon double bond in the place of a C4
hydroxyl; this may be a mechanism-based inactivator if the active site tyrosinate attacks
the double bond.
2.5.2 Inhibition of Dehydration
Compound 2 substitutes an amine group instead of a hydroxyl group on C4.
This compound could potentially be oxidized, but may not undergo imine formation
and certainly would not incur a negative charge on the nitrogen during the
dehydration step; however it retains the ability to form hydrogen-bonding interactions
with the enzyme. In fact, it is quite likely that the amine would be protonated at
physiological pH and therefore be even more drawn towards the tyrosinate ion.
Compound 3 would prevent the dehydration step from occurring, as it replaces the
proton on C5 with a fluorine substituent and would therefore not allow for proton
abstraction. However, it would mimic compound 14 in the reaction mechanism
catalyzed by RmlB. Compound 10 is similarly designed to allow for the oxidation step
but not the subsequent dehydration step, as there is no water available to eliminate
from the C6 position. Compounds 8 and 11, where the former possess an amine on the
-41~
C4 position and both contain fluorines on C6, are other variations intended to hinder
further reaction by preventing the dehydration/elimination step.
Compounds 4, 5 and 6 are inhibitors of the dehydration step designed to mimic
the geometry of an enolate, provided that the mechanism proceeds via an enolate
formation, as this reaction may be concerted, as mentioned earlier. Both Cl and Br are
suggested as oxygen atom mimetics, although the size of a chlorine atom is closer to
that of oxygen. The 6-deoxy form, compound 6, is also suggested, as it combines
features of both the dehydration and reduction step (Figure 15).
2.5.3 Inhibition of Reduction
Compounds 7 and 9 are end-product inhibitors using both fluorines and
hydrogens, respectively, on C6. Compound 9 is the actual product (compound 17 in
Figure 16) with hydrogens on C1 instead of a nucleotide group. As the enzyme
theoretically returns to its initial state after catalysis, these derivatives can mimic the
end product of the reaction and therefore should be accommodated by the active site.
2.5.4 Selection of Experimental Focus
While there were twelve different inhibitors designed originally, the
experimental approach of this thesis focused on the synthesis of three specific
compounds of the set. These three inhibitors, 9, 10 and 12, are proposed to resemble
molecules at different steps in the reaction mechanism, and compose a basic framework
for the initial evaluation of this strategy to inhibit RmlB.
Chapter 3. Experimental Methods
3.1 General
All reagents and solvents for syntheses and purification were obtained from
commercial sources (primarily Sigma-Aldrich and Toronto Research Chemicals) and
used without further purification. Reactions were followed by detecting compounds
using Hanessian's Cerium Molybdate stain on thin-layer chromatography plates (TLC).
1H and 13 C NMR spectra were recorded on the Bruker AVANCE-400 NMR
Spectrometer at the Department of Chemistry NMR Instrumentation Facility at MIT.
The 2D spectra (gCOSY, HETCOR and DEPT) were obtained using the Varian Inova500 NMR Spectrometer. For those 'H NMR spectra run in CDCl 3 , the internal standard
trimethylsilane was used as a reference point, while in other solvents, the residual
protons were used (CD 30D 6 4.89). For 13C NMR spectra run in CDCl3 , the peaks are
assigned relative to the residual carbons (6 77.23 for the central peak) while those run
in CD 3 0D are referenced to the central solvent signal at 6 49.15. The abbreviations
used are s, d, dd, t, q, and m, which are used to describe a singlet, doublet, doublet of
doublets, triplet, quartet and multiplet, respectively. All
J coupling values
are reported
in Hz, while chemical shifts are given in ppm.
3.2 Synthesis of 6-deoxy- 1,5-Anhydro-D-glucitol (10).
2,3-O~Acetyl-4,6-O-Benzylidene-1,5-Anhydro-D-glucitol
(19). 1,5-anhydro-D-
glucitol (18) (1 g, 6.1 mmol) was dissolved in DMF (5 mL). Benzaldehyde dimethyl
acetal (2.5 mL, 18.1 mmol) and a catalytic amount of p-TsOH (.116 g, .61 mmol) were
added and the reaction mixture was stirred at 40'C overnight. The reaction was
monitored by TLC and stopped after approximately 20 hours by the addition of
triethylamine (0.85 mL). The reaction was stirred for an additional 15 minutes and
allowed to cool to room temperature before being poured into H2 0 (120 mL). The
aqueous layer was washed with CHC13 (3 x 200 mL); the organic sections were
combined, dried using anhydrous Na 2 SO4 and concentrated, resulting in a yellow oil.
The mixture of compounds was carried on without further purification. The
benzylidene-protected glucitol (1.5 g, 6.1 mmol) was dissolved in pyridine (8 mL) and
acetic anhydride (20.0 mL) and heated at 80'C for 5 hours. The reaction mixture was
cooled to room temperature and poured into a mixture of H20 (150 mL) and EtOAc
(150 mL); after extraction with H20 (5 x 150 mL), the organic sections were dried
using anhydrous Na 2SO4 and concentrated. The desired product was purified by
column chromatography (5:1 hexane: ethyl acetate; Rf=0.3), yielding 1.6 g of a pure
white solid (80%yield). IH NMR (400 MHz, CDCl3): 6 7.43 (in, 2H), 7.35 (m, 3H),
5.50 (s, 1H), 5.34 (t, J=9.6 Hz, 1H), 5.05 (m, 1H), 4.33 (dd, J= 10.6 Hz, J=5 Hz, 1H),
4.13 (dd,J=1 1.2 Hz,J=5.6 Hz, 1H), 3.72 (t,J=10.2 Hz, 1H), 3.63 (t,J=9.6 Hz, 1H),
3.47 (m, 1H); 3.42 (t,J=8.2 Hz, IH), 2.08 (s, 3H), 2.05 (s, 3H);
13C
NMR (400 MHz,
CDCl3): 6 170.05, 169.92, 136.78, 128.95, 128.10, 125.99, 101.28, 78.63, 72.30,
71.31, 69.49, 68.44, 67.27, 20.75, 20.61.
2,3-O-Acetyl-4~O-Benzyl-1,5-Anhydro-D-glucitol (20). 2,3-~acetyl-4,6~Obenzylidene- 1,5-anhydro-D-glucitol (19) (400mg, 1.2 mmol) was dissolved into
BH3-THF (10 mL) and the reaction vessel was cooled to -10 C. A solution of .55 M
TMSOTf in CH 2 Cl 2 (1.1 mL) was added dropwise; the reaction was stirred for 1.5 hours
at -1
0C
until there was a significant change in the TLC. The reaction was poured into
ethyl acetate (50 mL), washed with H2 0 (3 x 50 mL), dried with Na2SO4 and
evaporated. The product was purified using column chromatography (2:1 hexane:
ethyl acetate; Rf=0.3) to yield 365 mg of a fluffy white solid in 90% overall yield. IH
NMR (400 MHz, CDCl 3 ): 6 7.25 (m, 5H), 5.18 (t,J=9.6 Hz, IH), 4.83 (m, 1H), 4.56 (d,
J=1.6 Hz, 2H), 3.99 (dd,J=11.2 Hz,J=5.6 Hz, 1H), 3.80 (dd,J=12 Hz,J=2 Hz, 1H),
3.63 (dd, J=12 Hz, J=4 Hz, 1H), 3.58 (t, J=9.6 Hz, 1H), 3.29 (m, 1H), 3.24 (t, J= 10.8
Hz, 1H), 1.97 (s, 1H), 1.95 (s, 3H), 1.90 (s, 3H); '-C NMR (400 MHz, CDCl 3 ): 6
170.40, 170.34, 137.73, 128.69, 128.15, 128.11, 79.93, 75.90, 75.73, 74.99, 69.76,
66.81, 61.77, 21.06, 20.92.
2,3-O-Acetyl-4-O-Benzyl-6-Deoxy-6-Iodo-1,5-Anhydro-D-glucitol
(21).
Triphenyl phosphine (233 mg, .89 mmol) and imidazole (120 mg, 1.8 mmol) were
added slowly to a stirring solution of 2,3-O-acetyl-4-O-benzyl- 1,5-anhydro-D-glucitol
(20) (200 mg, .60 mmol) in toluene (10 mL). Iodine (210 mg, .83 mmol) was
gradually added and the reaction stirred at room temperature for 3 hours. The reaction
mixture was then dissolved in EtOAc (2 x 25 mL) and washed with a saturated solution
of sodium sulfite (25 mL) followed by water (2 x 25 mL) and a saturated solution of
sodium chloride (1 x 25 mL). The organic section was dried with anhydrous Na 2 SO4
and evaporated and the residue purified using column chromatography (4:1 hexane:
ethyl acetate; Rf=0.3) to yield 185 mg (69%overall) of the slightly yellowish-white
solid. IH NMR (400 MHz, CDCl 3): 6 7.25 (m, 5H), 5.19 (s, 1H), 4.89 (m, IH), 4.65
(dd, J=15.6 Hz, J=4.4 Hz, 2H), 4.03 (dd, J= 11.2 Hz, J=5.6 Hz, I H), 3.47 (t,J=9.2 Hz,
1H), 3.42 (ddJ=11 HzJ=3 Hz, IH), 3.34 (m, 2H), 2.95 (m, IH), 1.96 (s, 3H), 1.92 (s,
3H); 13C NMR (400 MHz, CDCl 3): 6 170.34, 170.25, 137.58, 128.81, 128.34, 128.09,
79.78, 77.70, 75.71, 75.45, 69.60, 66.72, 21.11, 20.95, 7.12.
2,3-O-Acetyl-6-Deoxy-1,5-Anhydro-D-glucitol (22). 2,3-O-acetyl-4-O-benzyl,
6-deoxy-6-iodo- 1,5-anhydro-D-glucitol (21) (120 mg, .27 mmol) was dissolved in a
1:1 methanol/dioxane solution (10 mL). A small amount of triethylamine (100 pL)
was added and the solution was degassed with argon for 10 minutes. Activated Pd/C
(10% by wt.) was added (120 mg) and a H2 balloon was placed over the reaction,
which was then allowed to stir for 24 hours. At the end of 24 hours, the reaction was
filtered through Celite and evaporated to remove solvent and triethylamine. The
resulting solid was dissolved in degassed ethanol (24 mL), Pd/C (10% by wt.) was once
again added and the reaction stirred for an additional 24 hours. Once the mixture was
filtered through Celite and evaporated, the compound was purified using column
chromatography (2:1 hexane: ethyl acetate; Rf=0.3), resulting in 45 mg (7 1%) of a
crystalline white solid. IH NMR (400 MHz, CDCl3 ): 6 4.94 (m, 2H), 4.01 (m, 1H), 3.29
(m, 3H), 2.67 (s, 1H), 2.10 (s, 3H), 2.02 (s, 3H); '-C NMR (400 MHz, CDCL3): 6
172.23, 170.26, 77.55, 76.91, 74.71, 69.45, 66.81, 21.15, 20.95, 17.91.
6~Deoxy- 1,5-Anhydro-D-glucitol (10). The 2,3-O-acetyl-6-deoxy- 1,5-anhydro-Dglucitol (22) (45 mg, .19 mmol) was dissolved in a .014 M solution of sodium
methoxide in methanol (2 mL) and stirred for 2 hours. To stop the reaction, a small
amount of Amberlite Resin (H+) was added and the pH was monitored until it returned
to neutral pH was achieved. The resin was removed by filtration and the final product
was recovered in pure form after the solution was evaporated to dryness. The final
product was a white solid with a total yield of 20 mg (70%). IH NMR (400 MHz,
CD 3 OD): 6 3.80 (dd,J=11.2 Hz,J=5.2 Hz, 1H-), 3.42 (m, 1H), 3.21 (t,J=8.8 Hz, 1H),
3.17 (m, 1H), 3.12 (t,J=10.6 Hz, IH), 2.93 (t,J=9.2 Hz, 1H), 1.2 (d,J=6 Hz, 3H);
NMR (400 MHz, CDiOD): 6 79.77, 78.13,77.31,71.78, 71.08, 18.52.
-46-
13C
Chapter 4. Results and Discussion
In order to synthesize compounds 10, 9 and 12 in large quantity, it was
necessary to develop a synthesis that was reasonably short and efficient, based upon
existing methodology within the field of synthetic carbohydrate chemistry.
4.1 Retrosynthetic Analysis
As part of the initial project, the synthesis of twelve different inhibitors was
proposed, using a commercially available protected glucal as the starting material. For
the three compounds focused upon in this project, the original scheme (Figure 17)
involves a common intermediate and is quite analogous to the rationale described
below; the hydroxyl on the 6 position of the glucal would be removed, and the 4
position could be oxidized or methylenated to arrive at the desired products. The use of
a selectively protected glucal (Compound 25) as the starting material would make it
possible to add substituents to the C I position if desired, yet this double bond could also
be removed (to result in a 2-hydroxy) through the selective opening of a Danishevsky
epoxidation.4 0 However, it became obvious that this synthesis was not feasible; the cost
of the protected glucal was significantly prohibitive and synthesizing the protected
glucal from an unprotected glucal would have added five to six steps on to each
synthesis. As a result, a different synthetic approach was necessary.
The goal of synthesizing 12 different inhibitors was judged to be unrealistic for
an S.M. Thesis. Therefore, it was decided that a smaller number of inhibitors should be
chosen to focus efforts upon and attempt to synthesize these lead compounds in gram
quantities for animal antimicrobial experiments. The three inhibitors (_0, 9 and 1)
chosen were selected based on their synthetic accessibility, the possibility that they will
interact with RmlB at different stages of its catalytic mechanism and their close
-47-
structural relationship, which suggested they could be synthesized through a common
intermediate.
The strategy for the synthesis of these three inhibitors was to make them from a
common intermediate, Compound 30. Inhibitor 12 could conceivably be obtained
from intermediate 29 through a Wittig reaction, followed by removal of residual
protecting groups. Access to inhibitor 9 from this intermediate would proceed via a
similar deprotection step; synthesis of this key intermediate could be achieved by a
selective oxidation of the C4 position of another intermediate, Compound 30. This key
intermediate, which could link inhibitors 10, 9 and 12, consists of a selectively
protected version of inhibitor 10 that would enable the oxidation and methylenation of
C4 (Figure 18).
One option used commonly in carbohydrate synthesis to achieve a selective
protection is the benzylidene acetal protecting group. This group can simultaneously
protect the hydroxyls on the C4 and C6 positions and, in this case, would be more
useful than other conventional protecting groups (acetyl, benzyl, benzoyl or allyl) or
the other common acetal protecting group, the isopropylidene acetal, which may
protect the C4 and C6 positions but could protect any combination of adjacent
hydroxyls. In this case, if the C4 and C6 positions were protected with the benzylidene
acetal, the other two hydroxyl groups (on C2 and C3) could be protected with an
orthogonal protecting group.
Since the use of a benzyl protecting group would not be suitable as an
orthogonal protecting group as it is cleaved under the same conditions as the
benzylidene group, the typical possibilities include the allyl group, the benzoyl group
and the acetyl group. As allyl protecting groups are susceptible to cleavage under
conditions of catalytic hydrogenation, this would not be optimal, as that is the method
of removal of the benzyl group (resulting from benzylidene acetal cleavage). Between
the benzoyl and acetyl protecting groups, the acetyl group was chosen because of its
slight advantage in terms of ease in formation and cleavage.
With the benzylidene acetal in place and the other hydroxyls suitably protected,
the acetal could be regioselectively cleaved to protect the C4 hydroxyl group, leaving
the C6 hydroxyl available for removal. There are a number of ways that could be
employed to eliminate the hydroxyl group. However, since a hydrogenation must be
used to remove the benzyl group on the C4 position, it seemed as though the most
direct route would be to release the C6 hydroxyl with an iodine substituent that could
be simultaneously cleaved with a hydrogenation step. Finally, the benzylidene and
acetyl-group protected compound could be synthesized easily from 1,5-anhydro-Dglucitol. This latter compound is available from methyl a-D-glucopyranoside, an
inexpensive commercially available reagent, through a reduction of the anomeric
carbon (Figure 19).
4.2 Reduction ofMethyl a-D-Glucopyranoside
The proposed first step in the synthesis of the three proposed inhibitors involved
the reductive cleavage of methyl
-D-glucopyranoside. This step was attempted
according to the procedure published by Bennek et a, yet there were significant
differences in the results obtained and those reported. 4 ' The procedure involved an
aqueous workup that left the product in water, which was very difficult to fully
remove. Additionally, there appeared to be some side reactions with the silylating
reagents (trimethylsilyl trifluoromethansulfonate, triethylsilane and bis(trimethylsilyl)
trifluoroacetamide) used that resulted in the formation of a polymer type material that
was impossible to separate from the desired product. As a result, the yield was
unquantifiable. The material was carried forward through the benzylidene acetal
formation and acetylation steps, yet the contaminants remained unable to be separated
from the desired product despite multiple attempts at purification. In view of this
outcome, other options were considered; it did not appear that there was a less
problematic synthesis available for 1,5-anhydro-D-glucitol. However, a company
(Toronto Research Chemicals) was found that sold the 1,5-anhydro-D-glucitol at a
more reasonable price than Sigma-Aldrich, and after performing a cost analysis on the
materials consumed during the course of the reductive cleavage versus the cost of the
pure glucitol, it was decided to purchase the material instead of optimizing this step.
4.3 Benzylidene Acetal FormationandAcetyl Protection
The process of forming the benzylidene acetal according to the method of Ishii
etal resulted in high yield. The reaction is carried out with a catalytic amount of acid
and an excess of benzaldehyde dimethyl acetal, some of which remains unreacted upon
termination of the reaction. However, attempts to purify the desired compound at this
phase failed; the compound appeared to adhere strongly to the silica gel. Other
attempts to remove the benzyaldehyde dimethyl acetal were also unsuccessful: a
distillation of the crude oil and evaporation at a high temperature (60'C) to remove the
reagent resulted in decomposition of the desired product. Thus, it was decided to
acetylate the crude reaction mixture without purification, which appeared to be quite
successful. The acetylation was carried out in a standard fashion, and once the product
was acetylated, purification was facile.4 2 The yield of 75% over these two steps is
encouraging in the effort to synthesize larger quantities of material. The presence of
the desired product was verified by 1H NMR,
3
1 C
NMR and MS, yet a concern at this
stage was the inability to specifically assign the signals in the NMR spectra (Figures 21
and 22).
4.4 Assignment ofNMR Spectra and the Regioselective Ring Opening
Reaction
Perhaps the most complicated part of this synthesis was the regioselective
opening of the benzylidene acetal ring. There are a number of methods that could be
employed in the opening of this ring; the literature was explored for an option that had
high yields and specificity yet did not use extremely harsh conditions or particularly
hazardous chemicals. There were three possible procedures that were considered: one
used dibutylboron trifluoromethanesulfonate, a fairly hazardous chemical, another
required prolonged low temperature (-78'C) maintenance and the final one utilized
trimethylsilyl triflate, a hazardous chemical, yet less so compared to the dibutylboron
trifluoromethanesulfonate. Therefore, this latter procedure was chosen as the initial
method. The procedure reported by Jiang et al suggested that high levels of selectivity
were possible, but it was necessary to confirm the selectivity of the ring opening. 43 The
reaction was attempted according to the reported procedure, yet there was clearly some
formation of another isomer. It was unclear as to which isomer was formed
predominantly, but there seemed to be significant regioselectivity (about 9:1) when the
temperature of the reaction was approximately 00 C. When the reaction was repeated
and the temperature warmed up to almost 1 00C, the regioselectivity was greatly
decreased, and there was a mixture of the two isomers, in perhaps a 60:40 ratio.
Clearly, the regioselectivity of this reaction is temperature dependent. The optimal
temperature appears to be around ~1 0C, at which point the reaction proceeds to form
one isomer in a 97:3 yield.
The mechanism of this reaction is quite interesting. The selectivity in the ring
opening process is most likely due to steric reasons rather than electronic
considerations, as there is some of each isomer formed, and there is no clear electronic
~51~
difference between the two that favors opening towards one direction or another. The
silicon from the trimethylsilyl trifluoromethanesulfonate has a partial positive charge,
and, because of the bulky nature of the triflate and three methyl groups, the easiest
approach is most likely on the side of the C6 oxygen rather than the C4 oxygen, which
is shielded by the phenyl ring as well as the neighboring acetyl group on the C3
oxygen. This interaction results in bond cleavage between the C6 oxygen and the
benzyl methylene carbon, resulting in a proposed resonance stabilized transition state
structure (Figure 20). This proceeds to the final product, which consists of a free
hydroxyl on the C6 position and a 4-O-benzyl protected hydroxyl group. Despite this
rationale, it was clear that additional studies would be necessary to confirm that the
primary isomer formed was the sought-after product.
In order to determine which isomer was formed, the most logical choice was to
assign the peaks in the NMR spectra from the previous step and compare the changes to
the ring-opened product. In this vein, gCOSY, HETCOR and DEPT experiments were
carried out on both the pure 2,3-0-acetyl 4,6-O-benzylidene 1,5-anhydro-D-glucitol
(Compound 19) and the 2,3-0-acetyl 4-O-benzyl 1,5-anhydro-D-glucitol (Compound
20) (Figures 21, 22, 23, 24, 25, 26,27 and 28). In assigning the gCOSY spectra,
published IH NMR assignment for glucose was used as a reference. The peak at 5.34
ppm is attributed to H3. Using this as a reference, the remaining peaks were then
assigned. It was clear that the singlet at 5.49 ppm belonged to the proton from the
benzylidene acetal. The assignment of the peaks in the region between 3.00 and 5.50
ppm was significantly more complicated; the proton at 5.34 ppm splits with protons at
5.04 ppm, 4.12 ppm, 3.62 ppm and itself. While gCOSY spectra generally do not show
long distance coupling, it became clear that there exists some strong long range
coupling (coupling over several bonds) by protons in a similar region (Figure 23). It
would make sense, in terms of the spatial configuration of the proposed compound, that
the H3 proton would split with the H2, but it is unclear what the other two sources of
splitting could be. As H3 is an axial proton, protons in the general vicinity of the H3
area include H5 (also axial) and the axial proton off C1 (Hi axial). However, from the
HETCOR spectra (Figure 24), it was noted that the proton signals at 4.33 ppm and 3.70
ppm and the set at 4.12 ppm and 3.38 ppm arise from the two methylene groups (Cl
and C6). Therefore, it would appear that the set of protons at 4.12 ppm and 3.38 ppm
are close to H3 (in particular the H laxial proton), and are possibly the C1 protons. The
peak at 5.04 ppm is a complex multiplet, splitting with the protons at 5.34, 4.33, 4.12
and 3.38 ppm; because of its spatial proximity to H3 and the fact that it splits both the
protons from the C1 methylene group, this proton is assigned as H2. This leaves the
protons at 3.62 ppm and the multiplet at 3.47 ppm as unassigned. As the H5 proton is
more likely to be the proton splitting protons from both the C6 and C1 methylene
groups in addition to H2, the multiplet at 3.47 ppm is most likely H5, and the triplet at
3.62 ppm is then the proton off C4. The protons in the 7.30-7.50 ppm region are those
associated with the phenyl ring, while those at 2.04 and 2.06 ppm correspond to the
two acetyl groups. To confirm the assignments of these protons, these values were
compared to a similar molecule reported by Rye et a]44 The peak assignments appear
to be almost identical to those reported, with the exception of the anomeric position
which in the case of compound 19 is a methylene group. Additionally, the
13 C
NMR
can be assigned based upon the proton assignments, through the use of HETCOR
spectra. The 2 peaks at 170.22 and 170.09 ppm correspond to the carbonyl carbons
from the acetyl groups while the carbons from the phenyl ring demonstrate chemical
shifts at 137.00, 129.13, 128.29 and 126.19 ppm. The carbon from the acetal is
displayed at 101.46 ppm, while the peaks at 20.94 and 20.80 ppm correspond to the
methyl groups from the acetyl protecting groups. The other shifts are assigned based
upon the gCOSY assignments of peaks: C4 at 78.82 ppm, C3 at 72.51 ppm, C5 at 71.51
ppm, C2 at 69.69 ppm, C6 at 68.63 ppm, and C1 at 67.46 ppm.
Using these assignments, it is possible to analyze the specific changes in NMR
spectra between the benzylidene acetal protected compound 19 and the ring opened
product 20 to determine whether the benzyl group is in fact on the C4 oxygen or the
undesired C6 position. The IH NMR peaks in the aromatic region for this compound
are closer together than the precursor, while the acetal peak has disappeared as
expected (Figure 25). The triplet for H3 and the multiplet for H2 have shifted upfield
slightly to 5.23 and 4.90 ppm, respectively, as they are identifiable by their similar
splitting patterns in the gCOSY spectrum (Figure 27). The peaks at 4.62 ppm do not
split with any other protons, suggesting they belong to the methylene from the benzyl
group. There is now only one doublet of doublets in the range between 4.00 and 4.50
ppm; the doublet of doublets appearing at 4.05 ppm correspondsto one of the HI
protons. It is expected that one of the H6 protons shifts significantly, if this product is
the desired isomer, while the H4 proton would be expected to remain at nearly the
same chemical shift as the previous compound. The chemical shift of H4 is 3.69 ppm
compared to the 3.62 ppm from before, which is consistent with the idea that its
chemical environment would not change drastically due to the change from
benzylidene protected to benzyl protected. The remaining peaks for H5 and HI are
nearly the same as before, as is the H6 peak at 3.63 ppm. It is only a new doublet at
3.85 ppm that constitutes the major change, along with a peak at 2.25 ppm that
correspondsto a hydroxyl proton. This change in the second 1-16 proton signal further
confirms that the major product here is the 4-O-benzyl protected sugar, rather than the
6-0-benzyl protected compound. Therefore, it was determined that the reaction
conditions used to obtain compound 20 did indeed result in formation of the desired
isomer (Figure 29).
~54-
4.5 Iodination
There are numerous possible methods for the iodination of a 6-hydroxy group.
The primary consideration in the determination of the optimal experimental approach
was the selectivity of primary versus secondary alcohols, as it was desired to react only
the 4-O-benzyl product and not the small amount (less than 5%) of 6-O-benzyl
material, enabling complete purification of the desired product 21. As heating the
reaction would most likely encourage the reaction of the secondary alcohol, an
iodination reaction performed at room temperature was selected for use, despite this
procedure 4 5 having slightly lower yields than another procedure involving
cholorodiphenylphosphine and a higher temperature instead of triphenylphosphine. 46
The other option was to convert the free hydroxyl group into a triflate and then into a
halogen, however this would have added an unnecessary step and would most likely
have decreased yields.
The mechanism of the reaction in this case is relatively straightforward. The
triphenylphosphine interacts with the free hydroxyl group, and binds to the oxygen,
creating a good leaving group. This leaving group is displaced in an SN2-type fashion
by an iodine, resulting in the final product, Compound 21. While the reaction was
successful in this case, the yields can most likely be improved, as the conditions were
not optimized. However, the formation of the desired product has been confirmed by
IH and 13C NMR, and the assignment of the peaks from the synthetic precursors enables
the analysis of these spectra. The major changes between the ring-opened product and
this material are the loss of the hydroxyl proton peak and the shift of one of the H6
protons upfield (Figures 30 and 31).
~55-
4.6 Hydrogenation
The hydrogenation reaction was carried out according to a procedure by Takeo
et al,described as successful in removing an iodide. 4 7 This particular procedure was
chosen over others because it suggested conditions that did not involve the use of a Parr
apparatus (or any sort of high Hz pressure), and it removed the iodide cleanly, which
was thought to be a more difficult group to remove than the benzyl protecting group.
Originally, it was supposed that this step would simultaneously remove the
benzyl protecting group on the C4 position. However, when this reaction was
attempted using 1:1 methanol/dioxane as a solvent and a small amount of
triethylamine, it appeared that the iodide was removed, yet the benzyl group remained
on the C4 hydroxyl group. When the reaction mixture was filtered and subsequently
hydrogenated again in degassed ethanol without triethylamine, according to a method
reported by Gray et al,the removal of the benzyl group progressed efficiently. 48 At this
stage, the goal of hydrogenating both portions of the compound simultaneously was
still highly attractive, therefore the hydrogenation was attempted using normal ethanol
and no triethylamine, and the hydrogenation did not progress well. It seems that
removal of the iodide progresses neatly when triethylamine is present, which perhaps
neutralizes the acid formed in the reaction. However, the reaction may be solventdependent, as the two hydrogenation conditions for iodo and benzyl removal seem to be
optimized in slightly different solvent systems. It is also possible that the success of
hydrogenation is based upon the degassing of the solvent or the H2 pressure, as the
reactions went to completion only after ethanol degassing and two balloons of Hz
instead of one. Whatever the case, it appears that the slight differences in the nature of
these two hydrogenations necessitate setting up the reaction in two subsequent stages.
The yields are not quite as quantitative as described in the literature, but the overall
yield from these two steps is 71%; it may be possible to further optimize these steps. In
the end however, the desired product is obtained, as confirmed by NMR studies (Figures
32 and 33). In the 'H NMR spectrum, the benzyl protons have disappeared, while a
chemical shift corresponding to a methyl group has appeared (for C6), as well as a
hydroxyl proton peak (for C4). The acetyl groups are present in the same region, yet
the remaining protons have less distinct splitting patterns. The 13 C NMR spectrum
correspondsto the correct number of carbons in the appropriate regions.
4.7 FinalDeprotection and Peak A ssignment of Inhibitor 10
The deprotection of the two acetyl protecting groups is the final step in the
4
using a small amount of
synthesis. The method employed was a standard procedure,"
sodium methoxide in methanol, and an acidic resin to quench the reaction. The
advantage of this procedure is that there is no need to purify the final product, as this
compound is most likely too polar to purify using a silica gel column. The yields thus
far for this step have been lower (70%) than expected (quantitative), but this is most
likely due to the small scale of the reaction; it is difficult to measure such small
quantities of product accurately.
The IH and 13C NMR spectra were again studied using 2D NMR spectral
techniques for compound 10, as the assignment of the peaks was essential in
confirming the correct structure (Figures 34, 35, 36 and 37). The relative positions of
the sugar protons have changed, and although the exact chemical shift has changed
only slightly, there is an overall trend of movement upfield. The most downfield proton
is the doublet of doublets at 3.80 ppm, which correspondsto one of the HI protons,
similar to previous compounds. The multiplet at 3.43 ppm fits the usual splitting
pattern of the 12 proton, as it splits with the H I proton at 3.80 ppm along with other
protons in the 3.10-3.30 ppm range. The triplet, multiplet and triplet in the 3.10-3.30
ppm range overlap slightly, but it appears that the multiplet at 3.16 ppm is the H5
proton, as it has a similar coupling pattern as before. The remaining three triplets are
H 1, H3 and 114, and the question of which signals correspond to which protons is
revealed using the gCOSY spectrum. The peak at 3.12 ppm is clearly H1, as it splits
with the other HI proton and H2. The triplet at 3.21 ppm is most likely the H3 proton,
as it couples with the proton at 2.92 ppm and the H2 proton, and the triplet at 2.92
ppm couples with the proton at 3.21 ppm and H5. Spatially, it is more intuitive that
the H3 proton would couple with both H2 and H4 while the H4 proton would couple
with H3 and H5. Additionally, the doublet at 1.21 ppm corresponds to the methyl
group on C6; overall, the assignment of these peaks fits with the proposed product.
Due to time constraints, the finalized products 9 and 12 are not described here, yet they
are easily accessible from the successful synthesis of compound 10.
4.10 ConclusionsandFutureStudies
The long-term goals of this study are grounded in the mounting need for newer
therapeutics for TB. However, the primary goal of this project was to develop and
evaluate inhibitors for dTDP-D-glucose 4,6-dehydratase (RmlB). The efficacy of this
rational inhibitor design will be tested as these inhibitors move into biological tests; the
animal testing of these three inhibitors only touches the surface of the original goals of
this project. Encouraging results from the animal tests, to be performed by Professor
Eric Rubin at the Harvard University School of Public Health, could lead to the
continuation of this project. Completing the synthesis of the 12 proposed inhibitors as
well as expanding the compounds tested into a larger library would be the logical
progression, and it is the sincere hope of the author that this is accomplished.
Another possibility for the progressionof this project is to develop assays to
determine levels of enzymatic inhibition, cellular uptake and bioavailability of these
derivatives. As the reproduction of the bacteria in macrophages is a key factor in the
prolonged lifetime of the infection, an assay for studying the ability of these inhibitors
to treat infected macrophages would also be implemented. The initial set of inhibitors
will provide the framework for further library design and screening, perhaps using
high-throughput screening techniques.
There are significant long-range goals in the motivation for this study. From a
public health standpoint, TB is a serious killer: two million people each year die from
TB and a new person is infected with TB every second. It is estimated that in the next
twenty years almost one billion people will become infected, with over 200 million of
these people manifesting the debilitating symptoms of TB; clearly this is a serious health
problem that is not going away.
The dangers of a growing population of people infected with TB are largely
ignored because of the disease's prevalence in primarily developing nations. However,
people in the United States and the United Kingdom, particularly displaced or homeless
people in urban areas such as San Francisco and London, are at risk as well. Overall
risk is currently over 7% in the United States and at 13% in the UK; the rates of
infection in poorer countries are much higher. TB is a global health emergency, and
the most viable solution to this problem, especially considering the increase in multidrug resistant TB, is to develop newer treatments that can match or surpass current TB
drugs such as isoniazid and rifampin.4
The development of new therapies for TB is an exciting scientific problem, and
this study hopes to take a step in the direction of meeting this challenge.
References
1. World Health Organization. Global Tuberculosis Control: Surveillance, Planning, Financing.
WHO Report 2003. WHO/CDS/TB/2003.316. Geneva: 2003.
See http://www.who.int/gtb/publications/globrep/index.html
2.
Salyers, A and Dixie Whitt. Bacterial Pathogenesis: A Molecular Approach. Washington, D.C,
1994.
3.
World Health Organization. TB/HIV: A Clinical Manual. WHO/TB/96.200. Geneva 1996.
4.
World Health Organization. Tuberculosis: Fact Sheet No 104. Geneva: 2002.
See http://www.who.int/mcdiacentre/factshects/whol
04/en/
5.
World Health Organization. Western Pacific Region: Getting the Job Done Together. Geneva:
See http://www.wpro.who.int/pdf/WHOFlyer.pdf
6.
The World Bank Group. Country classification by income group according to 2001 GNI per
capita.
http://www.worldbank.org/data/databytopic/class.xls
7.
World Health Organization. Anti-tuberculosis Drug Resistance in the World. Report No. 2:
Prevalence and Trends. WHO/CDS/TB/2000.278. Geneva: 200.
See http://www.who.int/emc-documcnts/antimicrobial-resistance/docs/AntiTB/intro.pdf
8.
National Institute of Allergy and Infectious Diseases. Statement on World TB Day: March 24,
2003.
See http://www.niaid.nih.gov/newsroom/tbday03/statement.htm
9.
Draper, P. The outer parts of the mycobacterial envelope as permeability barriers. Front Biosci.,
3:D1253-61.: D1253-D1261, 1998.
10. Crick, D. C., Mahapatra, S., and Brennan, P. J. Biosynthesis of the arabinogalactan-peptidoglycan
complex of Mycobacterium tuberculosis. Glycobiology, 11: 107R- 118R, 2001.
11.
Raupach, B. and Kaufmann, S. H. Immune responses to intracellular bacteria.
Curr.Opin.Immunol., 13 417-428, 2001.
12.
Russell, D. G. Mycobacterium tuberculosis: here today, and here tomorrow. Nat.Rev.Mol.Cell Biol.,
2 569-577, 2001.
13.
Flynn, J. L. and Chan, J. Immunology of tuberculosis. Annu.Rev.Immunol., 19:93-129.: 93-129,
2001.
14.
FlynnJ. L. and Ernst,J. D. Immune responses in tuberculosis. Curr.Opin.Immunol., 12 432-436,
2000.
15.
Gordon, S. Alternative activation of macrophages. Nat.Rev.Immunol., 3: 23-35, 2003.
16.
Vanderlugt, C. L.and Miller, S. D. Epitope spreading in immune-mediated diseases: implications
for immunotherapy. Nat.Rev.Immunol., 2 85-95, 2002.
17.
van Crevel, R., Ottenhoff, T. H., and van der Meer, J. W. Innate immunity to Mycobacterium
tuberculosis. Clin.Microbiol.Rev., 15: 294-309, 2002.
18. Stenger, S. and Modlin, R. L. Control of Mvcobacterium tuberculosis through mammalian Toll-like
receptors. Curr.Opin.Immunol., 14: 452-457, 2002.
~60~
19.
Wallis, R. S. and Johnson,J. L.Adult tuberculosis in the 21st century: pathogenesis, clinical
features, and management. Curr.Opin.Pulm.Med., 7 124-132, 2001.
20.
Somoskovi, A., Parsons, L. M., and Salfinger, M. The molecular basis of resistance to isoniazid,
rifampin, and pyrazinamide in Mycobactcriun tuberculosis. Rcspir.Res., 2 164-168, 2001.
21.
Zhang, Y. and Amzcl, L. M. Tuberculosis drug targets. Curr.Drug Targets., 3: 131-154, 2002.
22.
Takayama, K., Armstrong, E. L., Kunugi, K. A., and Kilburn, J. 0. Inhibition by ethambutol of
mycolic acid transfer into the cell wall of Mycobacterium smegmatis. Antimicrob.Agents
Chemother., 16 240-242, 1979.
23.
Mikusova, K., Slayden, R. A., Besra, G. S., and Brennan, P. J. Biogenesis of the mycobacterial cell
wall and the site of action of ethambutol. Antimicrob.Agents Chemother., 39: 2484-2489, 1995.
24.
Alcaide, F., Pfyffer, G. E., and Tclenti, A. Role of embB in natural and acquired resistance to
ethambutol in mycobacteria. Antimicrob.Agcnts Chemother., 41: 2270-2273, 1997.
25.
Meier, A., Sander, P., Schaper, K.J., Scholz, M., and Bottger, E. C. Correlation of molecular
resistance mechanisms and phenotypic resistance levels in streptomycin-resistant Mycobacterium
tuberculosis. Antimicrob.Agents Chemother., 42 2452-2454, 1996.
26.
Mustafa, A. S. Development of new vaccines and diagnostic reagents against tuberculosis.
Mol.Immunol., 39: 113-119, 2002.
27.
McKinney, J. D. In vivo veritas: the search for TB drug targets goes live. Nat.Mcd., 0 1330-1333,
2000.
28. Sanders, D. A., Staines, A. G., McMahon, S. A., McNeil, M. R., Whitfield, C., and Naismith,J. H.
UDP-galactopyranose mutase has a novel structure and mechanism. Nat.Struct.Biol., : 858-863,
2001.
29.
Pan, F., Jackson, M., Ma, Y., and McNeil, M. Cell wall core galactofuran synthesis is essential for
growth of mycobacteria. J.Bacteriol., 183: 3991-3998, 2001.
30.
Zhang, Q. and Liu, H. w. Mechanistic Investigation of UDP-Galactopyranose Mutase from
Escherichia coli Using 2- and 3-Fluorinated UDP-Galactofuranoscas Probes. Journal of the
American Chemical Society, 123: 6756-6766, 2001.
31.
Daffe, M. The mycobacterial antigens 85 complex - from structure to function and beyond. Trends
Microbiol., 8: 438-440, 2000.
32.
Anderson, D. H., Harth, G., Horwitz, M. A., and Eisenberg, D. An interfacial mechanism and a class
of inhibitors inferred from two crystal structures of the Mycobacterium tuberculosis 30 kDa major
secretory protein (Antigen 85B), a mycolyl transferase. J.Mol.Biol., 307 671-681, 2001.
33. Jackson, M., Crick, D. C., and Brennan, P. J. Phosphatidylinositol is an essential phospholipid of
mycobacteria. J.Biol.Chem., 275. 30092-30099, 2000.
34.
Asselineau, C., Asselincau,J., Lancelle, G., and Lanecelle, M. A. The biosynthesis of mycolic acids by
Mycobacteria: current and alternative hypotheses. Prog.Lipid Res., 41: 501-523, 2002.
35.
Allard, S. T., Giraud, M. F., Whitfield, C., Graninger, M., Messner, P., and Naismith,J. H. The
crystal structure of dTDP-D-Glucose4,6-dehydratase (RmlB) from Salmonella enterica scrovar
Typhimurium, the second enzyme in the dTDP-1-rhannose pathway. J.Mol.Biol., 307 283-29 5,
2001.
36.
Giraud, M. F. and Naismith,J. H. The rhamnose pathway. Curr.Opin.Struct.Biol., 10 687-696,
2000.
-61-
37.
Ma, Y., Pan, F., and McNeil, M. Formation of dTDP-rhamnosc is essential for growth of
mycobacteria. J.Bacteriol., 184: 3392-3395, 2002.
38.
Allard, S. T., Beis, K., Giraud, M. F., Hegeman, A. D., GrossJ. W., Wilmouth, R. C., Whitfield, C.,
Graninger, M., Messner, P., Allen, A. G., Maskell, D. J., and NaismithJ. H. Toward a structural
understanding of the dehydratase mechanism. Structure.(Camb.), 10. 81-92, 2002.
39.
Chang, C. W., Chen, X. H., and Liu, H. w. CL)P-6-dceoxy-6,6-difluoro-D-glucose: A MechanismBased Inhibitor for CDP-D-glucose4,6-Dehydratase. Journal of the American Chemical Society,
120 9698-9699, 1998.
40.
Su, D. S., Meng, D., Bertinato, P., Balog, A., Sorensen, E.J., Danishefsky, S. J., Zheng, Y. H., Chou, T.
C., He, L., and Horwitz, S. B. Total synthesis of (-)-epothilone B: an extension of the Suzuki
coupling method and insights into structure-activity relationships of the cpothilones. Angcwandtc
Chemic, International Edition in English, 36' 757-759, 1997.
41.
Bennek, J. A. and Gray, G. R. An efficient synthesis of anhydroalditols and allylic-glycosides.
Journal of Organic Chemistry, 52 892-897, 1987.
42.
Rye, C. S. and Withers, S. G. Synthesis and Evaluation of Potential Inhibitors of Chondroitin AC
Lyase from Flavobacterium heparinum. Journal of Organic Chemistry, 67: 4505-4512, 2002.
43. Jiang, L. and Chan, T. H. Boranc/Bu2BOTf: a mild reagent for the regioselective reductive ring
opening of benzylidene acetals in carbohydrates. Tetrahedron Letters, 39: 355-358, 1998.
44. Rye, C. S. and Withers, S. G. Elucidation of the Mechanism of Polysaccharide Cleavage by
Chondroitin AC Lyase from Flavobacterium heparinum. Journal of the American Chemical Society,
124: 9756-9767, 2002.
45. Garegg, P. J. and Samuelsson, B. Novel reagent system for converting a hydroxy group into an iodo
group in carbohydrates with inversion of configuration. Journal of the Chemical Society, Chemical
Communications, 978-10, 1979.
46.
Classon, B., Liu, Z., and Samuelsson, B. New halogenation reagent systems useful for the mild onestep conversion of alcohols into iodides or bromides. Journal of Organic Chemistry, 53 61266130, 1988.
47.
Takeo, K. and Shinmitsu, K. A convenient synthesis of 6'-C-substituted b-maltose heptaacetates and
of panose. Carbohydrate Research, 133: 135-145, 1984.
48.
Gray, G. R. and Barker, R. Platinum-catalyzed hydrogenation and hydrogenolysis of acylated
pyranosyl halides, tetra-O-acctyl-2-hydroxy-D-glucal and tri-O-acetyl-D-glucal. Journal of
Organic Chemistry, 32 2764-2767, 1967.
49.
Agoston, K., Dobo, A., Rako, J., Kerckgyarto, J., and Szurmai, Z. Anomalous Zemplen dcacylation
reactions of a- and b-D-mannopyranosidc derivatives. Carbohydrate Research, 330. 183-190,
2001.
50.
Sisu, E., Sollogoub, M., Mallet, J. M., and Sinay, P. Cycloheptanic sugar mimetics, bridging the gap
in the homologous series of carbocyclicanalogues. Tetrahedron, 58: 10189-10196, 2002.
51.
Jorgensen, M., Iversen, E. H., and Madsen, R. A Convenient Route to Higher Sugars by 'lwo-Carbon
Chain Elongation Using Wittig/Dihydroxylation Reactions. Journal of Organic Chemistry, 66.
4625-4629, 2001.
-62-
Figures.
-63-
Country
India
China
Indonesia
Bangladesh
Nigeria
Pakistan
South Africa
Philippines
Russian Federation
Ethiopia
Kenya
DR Congo
Viet Nam
UR Tanzania
Brazil
Thailand
Zimbabwe
Cambodia
Myanmar
Uganda
Afghanistan
Zambia
Population
(thousands)
1,025,096
1,284,972
214,840
140,369
116,929
144,971
43,792
77,131
144,664
64,459
31,293
52,522
79,175
35,965
172,559
63,584
12,852
13,441
48,364
24,023
22,474
10,649
Estimated
Cases
Rate per
100,00
1,820,369
1,447,948
581,846
327,753
274,972
247,416
243,306
228,931
193,363
188,097
161,085
158,734
141,353
123,717
110,511
85,870
80,733
78,564
78,473
77,853
70,531
69,549
Figure 1. WHO statistics on global tuberculosis infection.
178
113
271
233
235
171
556
297
134
292
515
302
179
344
64
135
628
585
162
324
314
653
Figure 2. Scanning electron micrographs of Mycobacterium tuberculosis,courtesy of
http://www.bact.wisc.edu/Bact303/atlas (top) and
http://www.niaid.nih.gov/dir/labs/lhd/barry.htm (bottom).
Capsule-Like Material
Galactofuranose
Cell Membrane
Figure 3. Schematic representation of the cell wall of M. TB.
Early (innate)
lrnimature phagosome formed.
TNF-a and IFN-g secreted to
activate oxidative burst mechanisms.
TNF-a
IFN-g
11 RNI/RO
Activated infected macrophage
Late (acquired)
Mycobacterial antigens presented
by Cd I iolecules to T cells.
Lysis via granulysin is activated.
T cell
Graniulysin
y Cell
Lysed
Chronic
Granuloma contains mycobacteria
Infected
macrophage
Giant Foamy Cell
Lymphocyte
Macrophage
Figure 4. Stages of the immune response.
-67-
Infected
macrophiage
Lymphocyte
Granuloma
Granuloma Bursting
Activated infected
macrophage
/
4
LrN-g
J------------%]FNg
r
FFN-a
';'ITN-a
FNTN-
: TFN-a
'I'FN-aI:
L-argimne
Oxygen
Nitric Oxide
Superoxide
Fas/FasL
'eroxynitrite
Nitrogen Dioxide
Hydrogen Peroxide
S-nitrothiols
DUnitrosvl iron
con lexes
,e
Perforin
Granulysin -
Ilydroxyl and
Ferryl Radicals
Release of
Live Bacteria
Phagolysosone fusion
Acidification
Iron Restriction
Cell Lysis
Figure 5. The immune response against M. TB.
~68~
a
..HH
NHNH2
H2N
HOc)
Isonliazid
Stre ptomiycin
N
CONH 2
(N
Iyrazi nainide
jCO.
C2H5
H
N
HOH 2C
~~CH2 0H
N
H
112H5
Etharabutol
Rifanipin
Acetyl CoA
Oxaloacetate
/*e
malate synthase
Malate *
Glyoxylate
Citrate
i
vase
N
Isocitrate
Succinate
Fumarate
Other Fates
Succinate
Vhmm..Succinyl~CoA
93tc
Pyruvate + Glyceraldehyde 3-phosphate
phospICa ynthau
1 -deoxyxylulose-5-phosphate
0
UDP-GIcNAc
2-C- methyl-D-erythritol-4 -phosphate
Acetyl CoA
1DP-MurNAc
Isopentenyl Diphosphate
Malonyl CoA
UDP-MurNAc- Pentapeptide
Geranyl Diphosphate
Mycolvi ACP/CoA
-
WE,Z-Farnesyl Diphosphate
Decaprenyl Phosphate (C -P)
Lipid 11
-4
GlcNAc-P-P-C
Lipid II
v
Mycolyl-Man-P-C
Trehalose Mycolate
Rha-GlcNAc-P-P-C
0
(Galf) Rha-GcNAc-P-P-C
(Araf) (Galf) Rha-GlcNAc-P-P-C,
0
(Mycolyl) (Araf) (Galf) Rha-GIcNAc-P-
Peptidoglycan
C
Q
0
H1
H,
OH
:Z:
dTTP
HO-
HORmlA
0-dTDP
OP0 3
dTDP-D-Glucose
D-Glucose-1-Phosphate
H
H , Ziz
H
HORm1B
HI
HO
HI
0-dTDP
dTDP-D-Glucose
H I-
dTDP-4-keto-6-deoxy-D-glucose
H
OdTDP
Rm1C
0-dTDP
dTDP-L-lyxo-6-deoxy-4-hexulose
dTDP-4-keto-6-deoxy-D-glucose
CH3
H
H
-
-OdTDP
CH 3
RmID
IH
dTDP-L-Iyxo-6-deoxy-4-hexulose
dTDP-L-rhamnose
Figure 9. Biosynthesis of dTDP-L-rhamnose from glucose-i -phosphate.
Figure 10. The Structure of RmlB. Top: Ribbon diagram of entire enzyme. Bottom:
Detailed view of active site and substrates.
-73-
Figure 11. The active site of RmlB. Top left: Space filling view of the substrate and
cofactor within the active site. Top right: Parallel ring arrangement of NADH and
dTDP~D~glucose in the active site. Bottom: Proximity of important active site residues
to the substrate and cofactor.
~74~
NAD+
OH
H
H
OTyr1.67
O-
meH
Step 1: Oxidation
H.
H
d0
OH
O-dTDP
O-dTDP
Tvr167
13
14
OH
Step 2a: Deprotonation &
Enolate Formation
O,
0G
2
oH
7
-)-dTDP
OH
5~f
135
15
HO
0-
H
A 1340
H
Step 2b: Elimination of VVater
A
HO-HI
0-dTDP
O-dTDP
16
NAD-H
H.
H
H
H
O
0
Step 3: Reduction
O-dTDP
OH
OH
Tyr167
Tyr167
Figure 12. Proposed mechanism of action of RmlB.
17
4
OH
OH
HOHO
OH
H2N
HO
O
OH
2
1
CI
0
Br
HOO
OH
OH
5
4
H2N
0
OH
OH
7
6
CF 3
HO
HO
O
HO
10
OH
HO
HO
OH
9
CF 3
O
11
CH3
HO
8
CH 3
F OH
3
CI
OH
CF 3
O
OH
OH
O
HO
HO-
OH
Figure 13. The twelve putative inhibitors designed for RmlB.
H2 C
CH 3
O
HO0
12
OH
NAD+
>
H
O-
HOHa
H
OH
H
OH
H
Step
H
1: Oxidation
,'
OH
H
dH
O-TD
O-dTDP
Tyr167
13
14
H2 C
HO
CH3
O
O
HO
HO
OH
OH
1
12
Figure 14. Putative inhibitors designed againstthe oxidation step of the mechanism.
-77~
H
sOH
OH
H
Step 2a: Deprotonation &
Enolate Formation
OH
0
-
H
-
O-dTDP
OH
G14135
Gfu 13 5
15
.OH
CF3
H2N
H2 N
0
HO
F
OH
2
OH
OH
3
8
HO
0
H
A p134
Step 2b: Elimination of Wate r 0HA00
0
p13
HO-
O-d TDP
16
.OH
.0
Cl-
Br-
HO-
CI
HO
O
HO
OH
OH
4
OH
5
6
CF3
CH3
HO
HO
O0
OH
HO
-
OH
OH
10
Figure 15. Putative inhibitors designed against the dehydration step of the mechanism.
H
H
r"H
OH
,1-0
Step 3: ReductionH,
O-dTDP
O-dTDP
OH
OH
Tyr167
17
Tyr167
0
CF 3
0
CH 3
H 0))
HO
O
HO
OH
7
OH
9
Figure 16. Putative inhibitors designed against the reduction step of the mechanism.
k/
0
0
Is
0
00
/
0=
Figure 17. Retrosynthetic analysis using a protected glucal as the starting material.
-80-
Fths
0
0
-lcs stesatn
0
-U
O
O
0)
aeil08
0
6)
00
6)O
O
-a
-o
G)
O
-0
0
Oo
ersnhtcaayi sn
Figur0
o
00
0O
0
0
-0
co
I
1. C7 H6(OCH 3)2,
6 OH
HO
cat. p-TsOH
DMF, 400 C
0
O
3 2 OH
TMSOTf
2. Ac2 O, pyr
OAc
0H 2C 2 , -10 C
75%
800C
18
19
HO
P(Ph) 3 , 12
BnO
O0
AcO
HO
BH3
OAc
imidazole
toluene
Bn
CH 21
AcO
OAc
90%
69%
20
21
CH 3
AcO
O
OAc
0OH
PCC, I
AcO
1. H2 , Pd/C, Et3 N,
Dioxane/MeOH
2. H2, Pd/C, Etharo
H2C HC
3
4
OAc
PPh 3=CH 2 ,
KOtBu
AcO
OAc
71%
23
22
NaOMe,
MeOH
HO
HO
24
NaOMe,
MeOH
O OH3
OH3
OH
HO
NaOMe,
MeOH
H2C H3C
OH
HO
OH
70%
10
9
Figure 19. Synthesis of putative inhibitors 10, 9 and 12.
12
I
6
H
I
Figure 20. Suggested mechanism for the regioselective benzylidene ring opening
reaction.
I
CA)
W/-0
-84-
Figure 21. 'H Spectrum of Compound 19, 2,3-O-acetyl-4,6-O-benzylidene- 1,5anhydro-D-~g1ucito1.
eWV
0
0
0
0
0
dv.
0
0
4
0
4
Figure 22.
C.
IC
'9
0
40
0~,
0
0,
0
0
'a
IC
0
a-
IC
%J.
0
IC
13
-85-
C Spectrum of Compound 19.
......
.........
wmlp
L
_________
6
I,
2.5
*
3.0-
,m
*
U
*
S.
ew
U
*
a-
S
4.5--
mm
5..)
U
Us
S.
U
6
U
6.06.5-
It.
1.5
___
1.0
6.S
6.0
5.5
5.0
4.5
4.0
N
3.5
3.0
2.5
-A-
2.0
-kA
wU
3.43.6
3.8
4.0
U
ee
4.2
U
aS
4.4-
4.8
e
32.4
a
a
0
e
*
5.
5.6
5.4
5.2
5.0
4.8
Figure 23. gCOSY spectra of Compound 19.
4.6
4.4
4.2
4.0
3.8
3.6
3.4
F1
(ppm
3
4~0
6-
8130
Q
120
110
100
90
80
0
70
60
50
40
30
20
g
Actl
Fiue2.I
10
40
40
pcrmo
opud2,2,--ctl40bny015ahdoD
..
.....
. ...
....
I-.
Figure 26.
V
0
8 -
0
0
I-
C
13
---------
-89-
C spectrum of Compound 20.
A I N ft
oil
A.
A
-
A.
I -14_.A.
F2
(ppmT,
3.3.
3.5.
5
-r-
-
a
3
0
.
2
6..
6.5
-
a
U
7.5
7.5
7
6.5
5.5
1.0
5.0
4 5
4 0
3.5
3.0
.5
C
F23.4
F
3.6
08
4 .0-
4.24.44.
4.8-
9,
00
:3.0
5.2
5.2
5.0
4.8
4.6
4.4
Figure 27. gCOSY spectra of Compound 20.
-90-
4.2
4.0
3.8
3.6
3.4
-mam%.
...........
.. I
.Ai Li1_V
I
I
I
'I
Fl
I
(ppm4
3.0
9
3,5
0
0
4.0
N
4.5-
0
5.0-
9
0
5.5
6.0-
0
r0E
6.5-
O
7.0#
130
120
110
k
rITT1
100
,
90
80
F2
QC
0
I
I
II
1
70
(ppm)
60
50
40
I
30
Compound 19.
H
H
H 5.49
,H4. 3 3
H-
0
H
7.3-7.5
.H3.
H-
2.06
H 5.
H
0
H
4.12
H
/
H
\ 2.04
H
Compound 20.
H
H
j4.62
H 363
H
J. H 3.85
0
H
7.1-7.4
H3 .3 8
1.95
0
H
H
/\1.9o
H
H
Figure 29. 1H Peak assignments for compounds 19 and 20.
38
-93-
Figure 30. 1H spectrum of Compound 21, 2,3-0-acetyl-4-O-benzyl-6-deoxy-6-iodo1,5-anhydro-D-glucitol.
Figure 31.
0-
0
a,
0~
0%
0
0
4-
4
0
a-
'p-
0%0
13
000 NONNI, i
..........
........
......
....
M.Ma
-94-
C spectrum of Compound 21.
I
V
------------
Figure 32. 1H spectrum of Compound 22, 2,3~O~acetyl-6-deoxy-1,5-anhydro-Dglucitol.
Figure 33.
V
C
'9-
4C
C-
C
S
3
-96-
C spectrum of Compound 22.
..............
NO i MW
-97-
Figure 34. 1H spectrum of Compound 10, 6-deoxy- 1,5 -anhydro- D-glucitol.
40
4
Fiue3.1Csetu
4A6
L
VOM
fCmon
0
. . ........
1.
2.0
.4
3.0
a3
at .~
* a.
.9
a2
9
8
3.5-s0
'1~
4.
43
23
0
1.5
I'
kY~ ?
~~~ ~
A,I!
I.............................................
...........
.P
..................................................
......................................
..............
F2
-
(ppm)-
3.3.11!
<
'.1~
07'
I 'I...:
3.3
3.4..
C.-
Vi:)
3.5--
3
. 1
3.6'
to
C
'rt g
3.8-.
3.8 3,7
3.6
3.5
Figure 36. gCOSY spectrum of Compound 10.
3.i
3.3
3.2
3.1
3.0
2.9
i 11
11
I
I
Fl
(ppm)-0.0-
0.51.0I
6
1.52.0
0
U
2.5-
0
3.0-
Q
3.5C
U
|
4.0-
80
70
50
60
F2 (ppm)
19
0
40
30
20
0
0
Biography
The author was born on August 8, 1981 in Bridgeport, Connecticut. She grew
up in various cities in southern California and graduated from San Clemente High
School in 1998. She attended the Massachusetts Institute of Technology and received
an S.B. in Chemistry along with a minor in Brain and Cognitive Sciences in 2002.
While an undergraduate, she worked in the laboratory of Dr. Ann Graybiel on
neurochemical and behavioral modifications induced by dopaminergic agents. After
graduation, she joined the lab of Professor John Essigmann in order to pursue this
project, which she designed during the spring of her senior year. As part of her
master's, she accompanied MIT faculty to Bangkok, Thailand to teach a course on
Biotechnology and Engineering at the Chulabhorn Research Institute. Upon completion
of her master's degree in Molecular and Systems Toxicology from MIT in June 2003,
she plans on entering the University of California, Berkeley in pursuit of a doctorate in
Chemistry.
~101~